麻豆传媒

CHICAGO -- A randomized trial of chemoradiation (CRT) versus surgery for HPV-positive oropharyngeal cancer ended prematurely because of deaths in the surgery arm, but not before demonstrating potential for a nonsurgical approach to de-escalation, the principal investigator said here.

With follow-up for as long as 36 months (median of 17 months), no patient in the CRT arm had died or had disease progression. Patients randomized to transoral surgery had an overall survival (OS) of 89.1% and progression-free survival (PFS) of 83.5%.

Total grade 2-5 toxicity rates were similar between treatment groups, and only grade 2/3 anorexia (P=0.01) and dysgeusia (P=0.03) were higher in the CRT arm. Swallowing scores were similar between the two treatment arms, reported David A. Palma, MD, PhD, of Western University in Kingston, Ontario, during the American Society for Radiation Oncology (ASTRO) meeting.

"The primary radiation therapy approach achieved excellent oncologic outcomes in treatment de-escalation, with a moderate toxicity profile, and should be tested in phase III trials," said Palma. "Unfortunately, the primary surgical approach was associated with an upfront risk of treatment-related mortality and suboptimal progression-free survival."

The results, though incomplete, continued the ongoing search for safe and effective de-escalation strategies for HPV-positive oropharyngeal cancer, which accounts for the vast majority of new cases and has a more favorable prognosis as compared with head and neck cancers associated with smoking and alcohol consumption.

The optimal approach to managing HPV-related oropharyngeal cancer remains controversial, and current approaches and recommendations vary across the spectrum of intensity, said Palma. De-escalation is even more unsettled, as multiple strategies have been evaluated with varying degrees of success. Few clinical trials have directly compared de-escalation strategies, particularly surgical versus nonsurgical approaches.

The E3311 trial compared transoral surgery with standard versus reduced-intensity adjuvant radiation therapy. The two strategies resulted in similar 2-year PFS and functional outcomes. Following an initial report at the 2020 American Society of Clinical Oncology virtual meeting, invited discussant Hisham Mehanna, MD, of the University of Birmingham in England, said several studies of de-escalation strategies have tried to reduce trimodal therapy to bimodal by eliminating chemotherapy.

"The downsides, of course, are the survival efficacy, which, as compared to chemoradiotherapy, are not yet fully proven," said Mehanna. "The improvements in functional outcomes that are thought to happen are also not fully proven, and of course, a proportion of patients will need triple-modality treatment, which may have quality-of-life and functional detriments."

Phase III trials comparing primary surgery and primary radiation therapy are needed, but difficult to conduct, Mehanna continued. "The essential problem is that surgeons find it difficult to give up their surgery. I am a surgeon, so I can say that."

E3311 took a pragmatic approach and evaluated de-escalation of adjuvant radiation therapy, but patients in both arms still had surgery, Mehanna noted. Another study reported at ASTRO, took a similar approach, reducing the radiation dose in adjuvant CRT. That resulted in less toxicity with similar oncologic outcomes as compared with standard adjuvant CRT.

Palma reported findings from the abbreviated trial, which compared reduced-dose primary radiotherapy versus transoral surgery with or without reduced-dose adjuvant radiotherapy for intermediate-risk HPV-positive oropharyngeal cancer. The overarching goal of the trial design was to avoid trimodal treatment and its associated toxicities.

Eligible patients had pathologically proven HPV p16-positive oropharyngeal cancer (T1-2, N0-2, M0 disease) and were candidates for either treatment strategy. Patients assigned to primary radiotherapy received 60 Gy in 30 fractions for gross disease, 54 Gy in 30 fractions to high-risk subclinical areas, and 48 Gy in 30 fractions for low-risk nodal areas. Patients with multiple involved lymph nodes or one node >3 cm received concurrent weekly cisplatin.

Primary surgery included the primary site with neck dissection and mandatory ipsilateral external carotid artery ligation. Tracheostomy was recommended. Patients received adjuvant radiotherapy as needed: 50 Gy in 25 fractions for intermediate-risk features or 60 Gy in 30 fractions for positive margins or extracapsular extension.

The primary endpoint was OS for the two arms, with statistical considerations estimating a 2-year OS of 94% for both treatment arms. Secondary endpoints included OS and PFS comparisons between the two treatment arms. The trial required 140 patients total to differentiate an OS difference of 94% vs 84%.

The trial ended because of unacceptable toxicity after enrollment of 70 patients, 35 in each arm. Two patients randomized to surgery died after transoral robotic procedures, one of an oropharyngeal bleed and the other of cervical spine osteomyelitis caused by an infection of the cervical spin. The infection might have resulted from seeding from adjacent surgery and possibly propagated by adjuvant radiation therapy, said Palma.

The preliminary findings suggested "excellent" oncologic outcomes with the primary radiation approach. Statistical significance calculations were not performed because of early termination of the trial after 50% accrual, said Palma.

Swallowing assessment showed scores of 85-86 at 1 year. One patient in each treatment group required a percutaneous feeding tube, but no patient in either group had a feeding tube at 1 year.

Comment