Dr. Ezra E.W. Cohen speaks during the session “What’s Next in Immunotherapy for Head and Neck Cancer?”
“Head and neck squamous cell carcinoma (HNSCC) is the road to progress,” Dr. Cohen said, but more work is needed. “Tumor tissue is accessible, and we need to take advantage of that. Now is the time to start working together.”
The abstracts presented during this year’s session on immunotherapy were predominantly positive, albeit in small numbers of patients.
Pembrolizumab in HNSCC
Dr. Robert I. Haddad speaks during the session “What’s Next in Immunotherapy for Head and Neck Cancer?”
Other studies have shown a somewhat higher response to pembrolizumab in HPV-positive versus HPV-negative HNSCC, and pembrolizumab was associated with higher PD-L1 expression levels and GEP scores.1,2
Whole-exome sequencing data were analyzed in two subset cohorts from the KEYNOTE 012 trial: 34 patients who were PD-L1 positive (cohort B1) and 73 who were PD-L1 unselected (cohort B2). All patients and viral subgroups—HPV and Epstein-Barr virus (EBV)—were included.
A total of 73 patients were HPV and EBV negative; 34 patients were HPV or EBV positive. ML was significantly associated with response in both HPV-negative and HPV-positive subgroups, as well as in all patients in combined B1 and B2 cohorts (p = 0.0223). ML was associated with a response in the combined HPV-negative and EPV-negative patients (p = 0.0166) but not in those with an HPV-positive or EBV-positive status (p = 0.3851).
Similarly, increasing neoantigen load and positive tumor clonality status were significantly related to response in viral-negative patients (p = 0.0251 and p = 0.0157 for load and clonality, respectively) but not in HPV-positive or EBV-positive patients. Neoantigen load was similar in all patients (p = 0.0532), and tumor clonality was significantly different between patients with viral-negative and viral-positive tumors (p = 0.0164).
ML and GEP were weakly correlated and remained significant predictors in a multivariable model adjusted for each measure (ML p = 0.0349; GEP p = 0.0056). In ML- and GEP-associated PFS, a high ML and high GEP (of 125) had a better hazard ratio (HR) compared against a low ML or low GEP (of 57.5; HR 0.40, 95% CI [0.22, 0.71]).
“Both biomarkers may have utility in characterizing responses to anti–PD-1 therapies. These results were hypothesis generating for us” and warrant additional study, Dr. Haddad said.
Epacadostat Plus Pembrolizumab
Dr. Omid Hamid speaks during the session “What’s Next in Immunotherapy for Head and Neck Cancer?”
Of the 38 enrolled patients, 27 (71%) had a prior smoking history, 33 (87%) were men, 36 (95%) were white, 25 (66%) had prior cetuximab treatment, and 22 (58%) had a positive PD-L1 expression.
In the 38 efficacy-evaluable patients, 31 (82%) received one or two prior lines of treatment, and seven patients (18%) received three or more lines of treatment. Overall response rate (ORR) and disease control rate (DCR; combined complete response, partial response, and stable disease) for patients with one or two prior treatments were 39% (three patients had complete response, eight had partial response) and 65% (eight patients had stable disease), respectively, by RECIST. For patients with three or more lines of treatment, ORR and DCR were 14% (one patient had a partial response) and 43% (two patients had stable disease), respectively. Response was observed regardless of HPV or PD-L1 expression status.
“The safety profile was consistent with previously reported findings,” Dr. Hamid said. “The efficacy of the combination is also consistent with findings from other tumor types (melanoma, non–small cell lung cancer, renal cell carcinoma, urothelial carcinoma) and supports the phase III investigation of this combination in HNSCC.”
Dr. Cohen called the two presentations “encouraging” and said that, as researchers try to modify the tumor environment or affect the cellular milieu, “we may need even more combinations—maybe a combination of four agents.”
“We might even be able to start thinking about the ‘C’ word—curing patients with immunotherapy alone,” he said. “There are no more excuses not to do translational research. We owe it to ourselves, to our patients, and to the next generation to end cancer.”
Pembrolizumab and Chemoradiation for Locally Advanced HNSCC
Dr. Steven Francis Powell speaks during the session “What’s Next in Immunotherapy for Head and Neck Cancer?”
A total of 27 patients with oropharyngeal, hypopharyngeal, and laryngeal stage III/IVA or B HNSCC (any HPV status) who were eligible for cisplatin-based, definitive chemoradiation (CRT) were enrolled from November 2015 to August 2016 as part of a safety cohort.
Pembrolizumab was given at a fixed dose (200 mg intravenously) between 4 and 7 days before initiation of CRT, then every 3 weeks during CRT (i.e., two concomitant doses), and then after CRT for five additional doses. CRT consisted of 40 mg/m2 of cisplatin administered intravenously once weekly for six doses (maximum dose, 240 mg/m2); CRT was given concurrently with 2 Gy of radiation administered once daily for 35 fractions (total dose, 70 Gy). Safety was determined by the occurrence of dose-limiting adverse events (AEs) and immune-related AEs (irAEs) during CRT or pembrolizumab treatment. Efficacy was defined as the complete response rate at 150 days post-CRT completion. Secondary endpoints included progression-free survival, overall survival, locoregional control, distant-metastasis rate, and quality of life.
“The majority of patients had advanced disease and unresectable [tumors],” Dr. Powell said. “For the HPV cohort, there were a large number of unresectable primary tumors (with six at stage T4 in HPV-positive patients), and most had advanced nodal disease.” Fifteen HPV-positive patients (75%) and six HPV-negative patients (85%) reported tobacco use (> 10 pack-year history).
There were eight planned doses, and 21 patients (78%) completed all eight. Three discontinuations resulted from irAEs, and three discontinuations resulted after early neck resection (in two patients) or prolonged hospitalization.
“Radiation was well tolerated,” Dr. Powell said, and the patient characteristics were typical of what is seen at their center (mostly grade 3). The ORR at day 150 showed a complete response in 21 patients (78%; 95% CI [58%, 91%]): 17 patients in the HPV-positive group (85%; 95% CI [62%, 97%]) and four patients in the HPV-negative group (57%; CI [18%, 90%]).
“We realize our data cohort is small,” he said, but the results will add to the literature on biomarkers, timing, and potential combinations.
Dr. Ravindra Uppaluri speaks during the session “What’s Next in Immunotherapy for Head and Neck Cancer?”
A phase II trial has an accrual goal of 46 patients (still enrolling), but Ravindra Uppaluri, MD, PhD, of Brigham and Women’s Hospital and Dana-Farber Cancer Institute, described the protocol in which all patients received one dose of pembrolizumab 2 to 3 weeks before surgery, and only those with high-risk pathologic features were given postoperative adjuvant cisplatin and radiation followed by pembrolizumab (Abstract 6012).
Of the 25 patients currently enrolled, 24 (96%) have stage IV disease, 92% have cT4 disease, and 80% have cN+ disease, he said.
Preliminary observations include no serious drug-related AEs or unexpected surgical delays or complications and no locoregional recurrence/distant metastases (14 patients have a year or more of follow-up).
Dr. Antonio Jimeno speaks during the session “What’s Next in Immunotherapy for Head and Neck Cancer?”
In one patient, no response was seen at day 13, but a “big response” occurred at day 34. “It may be that surgery helps induce a response,” he said.
The incidence of HNSCC has increased approximately 50% from 2005 to 2015, and “cure rates are far from complete,” said discussant Antonio Jimeno, MD, PhD, of the University of Colorado Cancer Center. HPV-negative HNSCC outcomes have not significantly changed during the past decade, and the role of immune-directed modulation in definitive therapy “needs to be investigated.”
The addition of immune modulation “seems to be feasible in contemporary HNSCC cohorts,” he said. The two studies (albeit with early results) suggest a minimal added toxicity and favorable comparisons to institutional outcomes data, with pathologic indications of antitumor effects.
“This is a perfect storm of good things happening” in research, Dr. Jimeno said.
–Michelle Dalton, ELS