Comment on: Neoadjuvant chemotherapy approach fails to improve locally advanced cervical cancer survival
This trial by Sudeep Gupta, from Tata Memorial Centre in Mumbai, India, and colleagues represents the first prospective randomized trial comparing neoadjuvant chemotherapy (NACT) followed by radical surgery versus concurrent chemoradiation (CCRT) in women with locally advanced cervical cancer. The trial did not demonstrate the superiority of NACT plus surgery over CCRT in disease-free survival (DFS), and the authors concluded that CCRT remains the standard of care.
However, several points should be considered:
1. The trial suffered from a methodological bias linked to the accrual of 100 fewer patients than initially planned, thus potentially impacting the final results;
2. The neoadjuvant chemotherapy schedule used – carboplatin plus paclitaxel every 3 weeks – may not be the most active regiment in chemotherapy-naïve patients based on Japanese data suggesting that cisplatin should be used in chemotherapy-naïve patients instead of carboplatin;
3. The dose of radiotherapy used is slightly higher than the 80–90 Gy generally recommended by international guidelines: this may explain the higher 5-year DFS reached in the standard arm relative to that expected as per the initial hypothesis made by the authors themselves;
4. The toxicity of the CCRT arm is outstanding: 25% of patients were still experiencing vaginal distress 2 years after completion of treatment;
5. Forrest plot analysis suggests the equivalence of the two strategies in stage Ib–IIa cervical cancer with a clear advantage of CCRT in stage IIb disease;
6. The chosen primary endpoint, DFS, is not the most appropriate end point in a strategy trial and, in terms of overall survival, the treatment arms appeared similar.
Given all these consideration, before concluding that there is no place for NACT plus surgery in the treatment of locally advanced cervical cancer, I strongly suggest to wait for the results of the recently concluded EORTC trial addressing the same issue, the final results of which will be available by the end of 2018.