In the study’s intent-to-treat (ITT) population, the 3-year DFS rate was significantly higher among patients who received sequential chemoradiation than among patients who received concurrent chemoradiation or radiation alone.
“After an adjustment for baseline presence of lymph node metastasis, we found that sequential chemoradiation decreased, by 40% to 50%, the recurrence risk, compared with the other two groups,” said investigator, of Sun Yat-Sen University Cancer Center in Guangzhou, China.
Study rationale and details
“As we know, early-stage cervical cancer can be cured after treatment,” Dr. Huang said. “However, the risk of recurrence remains higher among patients with high-risk factors, including lymph node metastases, positive surgical margins, and parametrial invasion. Postoperative adjuvant radiation and chemotherapy may result in favorable survival for patients with high-risk factors.”
Although the efficacy of concurrent radiation with single-agent cisplatin has been demonstrated in primary treatment for local advanced cervical cancer, it has not been assessed as adjuvant treatment for early-stage cervical cancer in a randomized, controlled trial, Dr. Huang explained.
“Another question is whether patients with intermediate-risk factors could benefit from the concurrent chemotherapy [with] radiation,” he said. “So far, we don’t have enough evidence.”
To gain some insight, Dr. Huang and colleagues initiated the STARS trial. The trial enrolled patients with stage IB1-IIA2 cervical cancer and at least one adverse risk factor after radical hysterectomy. Patients had a median age of 48 years, and most had stage IB1 or IIA1 disease.
The patients were randomized 1:1:1 to receive radiation alone, concurrent chemoradiation, or sequential chemoradiation.
All patients received pelvic radiation at 45-50 Gy. Patients in the concurrent chemoradiation arm also received five doses of weekly cisplatin at 30-40 mg/m2.
Patients in the sequential chemoradiation arm received 60-75 mg/m2 of weekly cisplatin plus 135-175 mg/m2 of paclitaxel in 21-day cycles, with two cycles given before and two cycles given after radiotherapy.
The treatment groups were comparable with respect to histologic subtypes, lymphovascular invasion rates, parametrial or surgical margin involvement, deep stromal involvement, tumor grade, minimally invasive surgery rates, and neoadjuvant chemotherapy, Dr. Huang said. He added, however, that lymph node metastasis was lowest in the radiation-only arm.
In the ITT population, patients who received sequential chemoradiation had significantly better DFS compared with patients in the other treatment arms. The ITT population included 353 patients in the sequential chemoradiation arm, 345 patients in the concurrent chemoradiation arm, and 350 patients in the radiation-only arm.
The 3-year DFS rates were 90% in patients randomized to sequential chemoradiation, 85% in patients randomized to concurrent chemoradiation, and 82% in patients randomized to radiation alone (hazard ratios, 0.52 for sequential chemoradiation vs. radiation alone and 0.65 for sequential vs. concurrent chemoradiation; P = .03).
Subgroup analyses showed a DFS benefit with sequential chemoradiation versus radiation alone across histology types, tumor grades, and tumor size, Dr. Huang noted.
In the per-protocol population, sequential chemoradiation was associated with better DFS when compared with radiation alone. However, there were no significant differences between the sequential and concurrent chemoradiation arms or the concurrent chemoradiation and radiation-only arms.
The per-protocol population included 235 patients in the sequential chemoradiation arm, 190 patients in the concurrent chemoradiation arm, and 324 patients in the radiation-only arm.
In the per-protocol population, the 3-year DFS rates were 91% in patients randomized to sequential chemoradiation, 86% in patients randomized to concurrent chemoradiation, and 82% in patients randomized to radiation alone (HRs, 0.47 for sequential vs. radiation alone and 0.67 for sequential vs. concurrent; P = .026).
In the overall population, DFS was superior among patients with intermediate-risk versus high-risk factors. The 3-year DFS rates were 90% and 74%, respectively (P < .05).