CHICAGO — First-line therapy options for head and neck cancer should now include immunotherapy with pembrolizumab (Keytruda, Merck) given alone or in combination with chemotherapy (CT), says experts discussing results of the KEYNOTE-48 trial, and this represents a practice change, they note.
The new data are from the protocol-specified final analysis of KEYNOTE-48, which was presented here (abstract 6000) at the American Society for Clinical Oncology (ASCO) Annual Meeting.
Presenter Danny Rischin, MD, from the Peter MacCallum Cancer Centre, Melbourne, Australia, said: “These data support pembrolizumab plus platinum-based CT and pembrolizumab monotherapy as new first-line standard-of-care therapies for relapsed/metastatic head and neck squamous cell carcinoma (R/M HNSCC).”
The study showed superior overall survival (OS) with pembrolizumab alone or in combination with CT across patients with two separate PD-L1 (program death ligand 1) cut-offs compared with the control group, who were treated with CT alone. In addition, superior OS was seen for the total patient population receiving the combination.
However, pembrolizumab with or without CT did not provide significantly better objective response rates (ORR) and progression-free survival (PFS).
Discussant Vinita Noronha, MD, of the Tata Memorial Cancer Centre in Mumbai, India, agreed that the data are practice changing. However, she cautioned that several questions remained unanswered.
According to Noronha, the data do not provide guidance on which patient should receive pembrolizumab alone or in combination with CT. She asked why the response rate and PFS did not improve in patients. And she questioned whether there was a role for sequential therapy of CT followed by maintenance pembrolizumab or if all patients should receive the combination up front.
KEYNOTE-48 was a randomized, open-label study that enrolled 882 patients with R/M HNSCC, which was incurable with local therapies. Participants were a median age of 61 years, were predominantly men, and the majority were current or former smokers. Locoregional recurrence was seen in approximately 30% of patients.
Patients were randomized to three groups:
The first group received pembrolizumab monotherapy at 200 mg every 3 weeks for 35 cycles (n = 301);
A second group received pembrolizumab 200 mg every 3 weeks for 6 cycles along with standard CT (n = 281) followed by maintenance pembrolizumab 200 mg every 3 weeks for an additional 29 cycles; and
A third group received CT alone (n = 300).
CT provided was the accepted standard of care in this setting, the EXTREME regimen, which was a combination of cetuximab, carboplatin, or cisplatin, and 5-fluorouracil given at standard dosages every 3 weeks for 6 cycles followed by cetuximab 250 mg/m2 once a week as maintenance therapy.
The primary endpoint was OS in three patient populations: patients with tumors that had a PD-L1 combined positive score (CPS) ≥ 20 (found in 85% of patients), patients with tumors that had PD-L1 CPS ≥ 1 (in 44%), and the total patient population. CPS was defined as the percentage of PD-L1-positive cells (tumor cells, lymphocytes, and macrophages) compared with the total tumor cell numbers.
Interim analysis results were presented at the European Society of Medical Oncology 2018 annual meeting, as reported by Medscape Medical News.
Results of this final analysis had a data cut-off of February 25, 2019.
For patients with CPS ≥ 20, pembrolizumab and CT was superior to CT alone, with a hazard ratio (HR) of 0.6 (95% CI, 0.45 – 0.82; P = .0004). Median OS was 14.7 months for the combination versus 11.0 months for CT alone.
For patients with CPS ≥ 1, pembrolizumab and CT was also superior to CT alone. Median OS was 13.6 months for the combination versus 10.4 months for CT alone (HR, 0.65; 95% CI, 0.53 – 0.80; P < .0001).
However, ORR and PFS were considered not significant, as they did not meet the threshold of protocol-defined superiority for patients with tumors expressing CPS ≥ 1 or CPS ≥ 20 compared with CT alone.
The pembrolizumab plus CT combination also showed superior OS compared with CT alone for the total population of patients. With an HR of 0.72 (95% CI, 0.60 – 0.87), median OS was 13.0 months for the combination of pembrolizumab and CT and 10.7 months for CT alone.
Grade 3 to 5 adverse events were similar across the two groups (85.1% for pembrolizumab combination and 83.3% for CT alone). Immune- and infusion-related adverse events were seen in 4.7% and 0.7% of patients, respectively, receiving combination therapy. Corresponding rates for patients receiving CT were 8.4% and 2.1%. Discontinuation rates were also similar: 32.6% for patients receiving the combination and 27.5% for patients receiving CT alone.
Rischin also reported updated data for the comparison of pembrolizumab monotherapy and CT alone. “In this patient population, OS did not meet the criteria for significance at the superiority threshold [P = .0059],” he said.
However, OS with pembrolizumab monotherapy was superior to CT alone in patients with tumors expressing CPS ≥ 20 (HR, 0.58; 95% CI: 0.44 – 0.78) and CPS ≥ 1 (HR, 0.74; 95% CI, 0.61 – 0.90).
Although responses with immunotherapy were more durable (median duration of response: 22.6 vs 4.5 months), response rates were lower compared with CT alone (16.9% vs 36.0%).
Implications for Clinical Practice
In discussing these data, Noronha provided a simplified take on the data.
For pembrolizumab monotherapy, ORR and PFS showed no benefits compared with CT, she noted. Toxicity was similar to or less than CT. Although pembrolizumab monotherapy was superior to CT for patients with CPS ≥ 20 and ≥ 1, it was not superior for the total population.
For the pembrolizumab and CT combination, ORR was similar to CT, duration of response was longer, PFS was again not improved in any of the patient populations, but OS was improved in all patient populations. “These data are practice changing,” she said.
She specifically called out the higher ORR seen with CT compared with pembrolizumab monotherapy. It is important to keep this in mind, especially when a response is needed in patients, she observed.
Noronha provided an algorithm for treating patients with HNSCC. For patients with a disease-free interval (DFI) of 6 months or less (platinum-refractory patients), monotherapy with immunotherapy is the treatment of choice.
If patients with a DFI of more than 6 months have a good performance status, controlled comorbidities, are eligible for platinum therapy, and can afford therapy (reimbursable), therapy is likely to be dictated by disease severity. She recommended pembrolizumab in combination with CT for patients with severe disease.
However, for patients with mild-to-moderate disease, PD-L1 levels should determine treatment approach. For patients with tumors showing CPS ≥ 20, pembrolizumab monotherapy may be appropriate; those with CPS ≥ 1 may be candidates for pembrolizumab monotherapy or pembrolizumab combination therapy. However, she added that “patients with CPS 0 should be given the EXTREME or a similar CT regimen or the pembrolizumab combination.”
New First-Line Option
Described as the “most important” study in the Highlights of the Day session the following morning, Francis P. Worden, MD, of the University of Michigan Rogel Cancer Center, Ann Arbor, predicts that the pembrolizumab plus CT combination will replace the EXTREME regimen as first-line therapy in HNSCC.
Both Noronha and Worden suggested that superior OS seen with the pembrolizumab combination in patients with CPS ≥ 1 and in the total patient population may have been driven by the subset of patients with CPS ≥ 20. Worden said it remains to be seen whether patients with CPS ≥ 1 to 19 and CPS 0 benefit from the combination of pembrolizumab and CT.
As a follow-up on Noronha’s recommendations, Worden suggested that as clinicians await approval for this new first-line indication for pembrolizumab, they should begin to undertake PD-L1 CPS testing and carefully evaluate patients with respect to disease progression, rate of disease progression, and bulk of disease.
However, he suggested that more is not always better. “We may be doing a disservice to our patients who have a poor performance status and CPS < 1 or 0 who may be better served by CT followed by immunotherapy as per the current indication,” he said.
Noting the difficulties associated with administering the EXTREME regimen (eg, 5-fluorouracil requires a central line), Worden pointed out that it is unknown if similar results would be seen if pembrolizumab was given in combination with another CT regimen.
“However, exposure to immunotherapy may have benefits for second-line therapies,” Worden said.
Rischin has reported receiving research funding from Amgen, Bristol-Myers Squibb, Genentech/Roche, GSK, Merck, and Regeneron. Noronha has reported no relevant financial relationships. Worden has reported receiving honoraria, consultancy fees, research funding, and/or travel assistance from Bayer, BMS, Cue Biopharmaceuticals, Eisai, Fusion Pharmaceuticals, LOXO, MSD, and Pfizer.
ASCO 2019. Presented May 31, 2019. Abstract 6000.