De-Intensified Therapy Misses Mark in HPV-Positive Oropharyngeal Cancer

Two strategies to de-escalate treatment for human papillomavirus (HPV)-positive oropharyngeal cancer failed to show noninferiority versus standard care, which “reset the bar” for future trials of de-escalation, a randomized trial showed.

Standard treatment with a 70-Gy dose of radiation therapy (RT) plus cisplatin led to a 2-year progression-free survival (PFS) of 98.1%, as compared to 88.6% and 90.3% with de-escalated strategies with 60 Gy of RT plus cisplatin or nivolumab (Opdivo). Not a single patient who received standard therapy had locoregional recurrence during the first 2 years of follow-up. Overall survival (OS) at 2 years was excellent in all three arms, topping out at 99% with standard therapy.

The results do not slam the door on de-intensified RT for HPV-positive oropharyngeal cancer, but better selection criteria are needed, said Sue Yom, MD, PhD, of the University of California San Francisco, at the American Society for Radiation Oncology (ASTRO) meeting in Washington.

“De-intensification studies, going forward, must evolve to be more selective and more effective if they are going to be more competitive in a more challenging population,” said Yom. “We also know that, while de-intensified radiation regimens worked well in this trial and have worked well in other trials, in the large majority of the patients, the eligibility criteria used in this trial are not selective enough. We need further enhancement of selection mechanisms, whether at entry or on treatment.”

“The bottom line is that the highest rate of cure now documented in the national head and neck cancer trial literature comes out of this study,” she added. “We can say that phase II trials should not be considered valid unless they are tested against a contemporary standard of care. These results set a new benchmark for PFS expectations in this population, and further de-intensification trials will now be held to this very high standard going forward.”

Calling the 2-year PFS of 98% “exceptional,” ASTRO invited discussant Sewit Teckie, MD, of NYC Health and Hospitals, said future de-escalation strategies should be compared against that high standard.

“We had previously accepted that in phase II trials, a progression-free survival rate of 90% was acceptable,” said Teckie. “I would argue that that number is no longer acceptable. The standard of care should remain 70 Gy and cisplatin. There is still a debate between [every 3 weeks] dosing, two cycles of high-dose cisplatin versus weekly dosing. Regardless, this is the standard of care. This is the standard of care in my own practice, 70 Gray over 7 weeks with weekly cisplatin.”

“Just as we learned that cetuximab (Erbitux) is inferior to cisplatin, we are now seeing that de-escalation based on our typical criteria is inferior to standard of care,” she added. “The bar for future de-intensification is reset, and it is high.”

The issue of de-intensification has come to the forefront of treatment for head and neck cancer with the evolution of HPV infection as the predominant cause of the cancer. About 70% of new oropharynx cancers are HPV-positive, which is associated with more favorable outcomes as compared with historical epidemiology tied to smoking and alcohol use, Yom noted. Increasingly, late toxicities associated with chemoradiation have become a major concern.

The randomized, phase II NRG-HN002 trial showed that a 60-Gy dose of RT with concurrent cisplatin met the existing acceptability benchmark for continued investigation in the phase II/III NRG-HN005, which compared two de-escalation strategies against standard care. Investigators in the multicenter trial enrolled patients with newly diagnosed HPV p16-positive oropharyngeal cancer, T1-T2N1M0 or T3N0-N1M0, and ≤10 pack-year history of smoking.

Patients were randomized to a control arm of 70 Gy RT over 6 weeks plus two cycles of cisplatin, 60 Gy RT over 6 weeks plus cisplatin, or 60 Gy RT over 5 weeks plus six cycles of nivolumab. During the phase II component of the trial, a futility analysis was triggered when 50% of PFS events (11 of 22) had occurred. A hazard ratio of

The trial would continue to phase III if either experimental arm survived two planned futility analyses. The primary endpoint for phase II was PFS noninferiority, and phase III co-primary endpoints were PFS noninferiority and superior quality of life (QoL).

Data analysis included 380 patients who had a median age of 60. Men accounted for 90.6% of the population, 87.5% were white, 84.9% had stage I disease, and 79.4% were never smokers.

At the first futility analysis after 1.1 years of follow up, two PFS events had occurred in the control arm and nine in the 60 Gy/cisplatin arm, which translated into an estimated hazard ratio of 4.34. Randomization was suspended for 3 months to discontinue the 60 Gy/cisplatin arm, and then enrollment resumed in the other two groups. A second futility analysis was triggered after 1.7 years. Two events occurred in the control arm and nine in the nivolumab arm, producing an estimated hazard ratio of 4.51.

Data analysis after a median follow-up of 2.2 years showed that neither experimental arm met criteria for PFS noninferiority. The 2-year locoregional failure rates were 0% in the control arm, 6.5% with 60 Gy/cisplatin, and 5.0% with 60 Gy/nivolumab. The 2-year OS was 99.0% with standard care, 98.0% with 60 Gy/cisplatin, and 96.1% with 60 Gy/nivolumab.