-- Postop ICI plus CRT offers much-needed boost to standard-of-care treatment for high-risk disease
CHICAGO - The postoperative addition of nivolumab (Opdivo) to chemoradiotherapy (CRT) extended disease-free survival (DFS) in certain patients with locally advanced head and neck squamous cell carcinoma (HNSCC), according to the NIVOPOSTOP trial.
Patients in the phase III study were at high risk of recurrence, and were not selected for PD-L1 expression. At a median of 30.3 months, patients randomized to nivolumab plus CRT had a 24% improvement in DFS (stratified HR 0.76, P=0.034, 95% CI 0.60-0.98) versus patients who got CRT alone. DFS was the primary endpoint, according to Jean Bourhis, MD, PhD, of Lausanne University Hospital in Lausanne, Switzerland.
Add-on nivolumab represents the first improvement to the standard of care (SoC), which is adjuvant high-dose cisplatin plus RT only, for locally advanced HNSCC in more than 2 decades, said Bourhis at a press briefing at the American Society of Clinical Oncology (ASCO) annual meeting.
He noted that postoperative nivolumab-CRT "could be proposed as a new standard treatment."
Up to 45% of patients with locally advanced HNSCC will experience recurrence following standard adjuvant cisplatin and RT, "defining an unmet clinical need," Bourhis said.
PD-1 inhibitors have been established as the SoC for recurrent or metastatic HNSCC. The KEYNOTE-689 trial demonstrated an improvement in event-free survival (EFS) with the addition of neoadjuvant pembrolizumab (Keytruda) followed by adjuvant pembrolizumab to SoC in patients with locally advanced resectable HNSCC. NIVOPOSTOP is the first study to test the efficacy of adjuvant PD-1 inhibition in a population similar to that studied in KENOTE-689.
The current data suggest that moving immune checkpoint inhibition (ICI) to the low-tumor burden setting postoperatively in locally advanced HNSCC is more effective than in the pre- or perioperative setting, said Bourhis. He noted several studies that produced suboptimal results when ICI was used with bulky tumors in place.
NIVOPOSTOP included 680 patients, ages <75, with stage III or IV SCC of the oral cavity, oropharynx, larynx, or hypopharynx. They had complete macroscopic surgical resection and high-risk pathological features of relapse. The latter was defined as extracapsular extension (ECE) in cervical nodes, microscopically positive tumor margins, involvement of four cervical nodes without ECE, and multiple perineural invasions.
Patients were randomized 1:1 to CRT (66 total Gy of radiation with systemic cisplatin) every 3 weeks for 3 cycles or an initial 240-mg dose of nivolumab followed by 360 mg nivolumab in combination with CRT every 3 weeks for 3 cycles, then 6 cycles of 480 mg of nivolumab alone every 4 weeks.
The rate of high-risk features was similar between the two arms. More than 80% of patients were current or former smokers, with a median number 40 packs per year in both arms. The most common tumor site was the oral cavity (58% of each arm). Overall, about 83% had stage IVA or IVB disease "suggesting very advanced local disease in most patients," he said.
The 3-year DFS rates were 63.1% in the nivolumab arm versus 52.5% in the control arm. The benefit of nivolumab on DFS was present in all-comers, and centrally assessed PD-L1 expression using the combined positive score was not a strong predictor of nivolumab efficacy.
"Most of the benefit came from improvement in locoregional control," said Bourhis. By year 3, the cumulative incidence of locoregional relapses alone were 13% versus 20%, respectively (stratified sub-HR 0.63, 95% CI 0.42-0.94).
Nivolumab did not compromise compliance with RT as more than 90% in each arm received a 66-Gy dose. Compliance to nivolumab before, during, and after CRT was high, with a median number of 10 nivolumab cycles, as per protocol. The median number of maintenance nivolumab cycles was 6.
Most adverse events (AEs) observed in the trial were grade 1 or 2, and most of the AEs were directly related to CRT, such as stomatitis and radiation skin injury, according to Bourhis. There was a slight increase in renal toxicity (24% vs 15%) and a more pronounced increase in thyroid disorders (20% vs 2%) in the nivolumab arm, as expected.
Overall survival (OS) data from NIVOPOSTOP are immature. At data cutoff, 158 patients died. Interim OS results favor nivolumab: OS probability at 3 years was 74.2% in the nivolumab arm and 67.8% in the control arm but OS could not be formally tested because the prespecified number of deaths has not been reached.
Given the efficacy of ICI in both KEYNOTE-689 (neoadjuvant plus adjuvant) and NIVOPOSTOP (adjuvant), the optimal timing for immunotherapy administration in patients with locally advanced HNSCC remains unclear, noted ASCO expert Glenn J. Hanna, MD, of Dana-Farber Cancer Institute in Boston.
"What is the right sequence of immunotherapy," asked Hanna. "Do you give immunotherapy first or do we wait and do it in the adjuvant setting?" The improvements in EFS/DFS were comparable between KEYNOTE-689 and NIVOPOSTOP across a PD-L1-unselected high-risk population, he noted, adding that patient selection for preoperative versus postoperative ICI therapy needs to be explored.
The benefit to pembrolizumab in KEYNOTE-689 was mainly a relative reduction in distant metastatic disease. "Perhaps giving immunotherapy upfront mitigates micrometastatic potential" whereas the impact to adjuvant nivolumab appeared to be locoregional, Hanna added.