Faster Radiation More Effective in Head and Neck Cancer

COPENHAGEN, Denmark—Radiation in 6 fractions per week is significantly better than the same dose given on a more leisurely 5-fractions-per-week schedule for treating squamous-cell head and neck cancer, according to investigators from the Danish Head and Neck Cancer Study Group (DAHANCA).

Jens Overgaard, MD, of Aarhus University Hospital, Aarhus, Denmark, Hanne Sand Hansen, MD, of Rigshospitalet, Finsen Centre, Copenhagen, and their colleagues reported data from a large randomized controlled trial comparing the two schedules in the September 20 issue of The Lancet (362:933-940, 2003). They also reported the results at the ECCO 12 meeting in Copenhagen.

"Six fractions per week is significantly more effective than 5 fractions per week with regard to 5-year locoregional control, primary tumor control, voice preservation in laryngeal cancer, and disease-specific survival," Dr. Hansen told ONI. The 6-fractions-per-week schedule is now standard treatment for head and neck cancer in Denmark.

The 6/wk schedule did not improve outcomes for patients with N2 or N3 tumors, and Dr. Hansen said that the research group is now planning a combined approach with radiation and surgery for such patients.

The trial randomized 1,476 patients with invasive squamous-cell carcinomas of the head and neck to 5 (n = 726) or 6 (n = 750) fractions per week of radiotherapy. Radiotherapy was given with 4 to 6 MV photons to standard fields including the primary tumor and involved lymph nodes. The minimum tumor dose was 62 to 68 Gy, with larger tumors receiving larger doses.

The idea behind the accelerated 6-fractions-per-week schedule is that it might improve tumor control by reducing the development of resistant clones, which are thought to arise through radiation-induced accelerated proliferation of clonogenic tumor cells.

The trial included two subprotocols: DAHANCA 6, including all glottic carcinomas, and DAHANCA 7, including tumors of the supraglottic larynx, pharynx, and oral cavity. Patients with distant metastases were excluded.

All patients except those with glottic cancer also received nimorazole as a radiosensitizer given at 1,200 mg/m2 orally with the first 30 radiation treatments, to a total dose of about 36 g/m2. Nimorazole was given 90 minutes before radiation.

The investigators noted that in other countries many of these patients might be treated initially by surgery, but the Danish standard is primary radiotherapy because it is more organ conserving and leaves more patients with voice intact.

The primary study endpoint was locoregional control, defined as "complete and persistent disappearance of disease in the primary tumor and regional lymph nodes after radiotherapy." The study was powered to detect an improvement in locoregional control from 55% to 65%.

Study Results

The accelerated schedule produced better locoregional control than a conventional schedule with identical dose and fractionation (odds ratio 0.66) (see Table). The accelerated schedule also significantly improved disease-specific survival (odds radio 0.71) but did not improve overall survival.

Analysis of tumor-site (T site) and nodal-site (N site) failures showed that the entire benefit was due to improved disease control at the primary tumor site. This was also demonstrated by significantly higher preservation of the larynx and voice in patients with laryngeal cancer (80% vs 68%, P = .007).

There was a significant increase in acute confluent mucositis with the 6- vs 5-fractions-per-week schedule (53% vs 33%, P < .0001), and the mucositis persisted longer in the 6/wk patients. "The mucositis subsides spontaneously. We treat it with local analgesics and with antibiotics and antiviral medication if necessary," Dr. Hansen said. All mucositis resolved within 3 months of the start of treatment.

"The window of opportunity for the benefit of acceleration is narrow, and with the applied radiation technique, a 1-week reduction seems to be the optimum balance between improved tumor control and avoidance of excess late morbidity," the investigators said in The Lancet.

Patients with large nodal burden (N2-N3) did not benefit from the accelerated schedule, regardless of the tumor site. Those with no or small nodal involvement had substantially better locoregional control with the 6/wk schedule. The reduction in treatment time was also more beneficial to patients with moderately to well-differentiated tumors than to those with poorly differentiated tumors.

The researchers suggest that the mechanism of repopulation in squamous-cell head and neck carcinomas might resemble the repair mechanisms of normal mucosa: The primary tumor might respond to the trauma of irradiation with a typical epithelial attempt at quick recovery. "Furthermore, the reaction might be controlled by signaling from the surrounding normal mucosa, and the response is, therefore, seen only in the T site and not in the nodal metastases," they said.