Adjuvant and neo-adjuvant chemotherapy in non-small cell carcinoma. Discussion

Background : There have been many trials investigating the influence of chemotherapy on survival following surgery or radiotherapy in non-small cell lung cancer. Few have been large enough to detect the sort of differences likely to emerge, given the limited efficacy of chemotherapy. The technique of meta-analysis using individual patient data can allow worthwhile conclusions to be derived from the contradictory data generated by multiple small trials. This paper summarizes published data from randomized trials testing : (1) adjuvant chemotherapy following 'curative' surgery ; (2) neoadjuvant chemotherapy prior to surgery in cases of borderline operability ; (3) chemotherapy in inoperable, but still localized, disease where standard therapy would be radical radiotherapy. Conclusion : All the indications are that combinations including cisplatin confer a small, but real, prolongation of survival. However, meta-analysis is not a substitute for individual trials large enough to detect clinically important differences in survival. Other worthwhile endpoints like symptom control, quality of life and cost cannot be addressed in a meta-analysis. Large, well designed and executed randomized trials are still urgently needed if more time and money is not going to be wasted in the search for better treatments in lung cancer.