EGFR-TKIs are the most promising development in the treatment of advanced NSCLC. Clinical trials comparing TKIs and placebos have produced controversial results. It seems that not all NSCLC patients benefit from these drugs. Different selection criteria for patients will produce different results.
The ORR of TKIs was higher in EGFR mutation selected groups than the other two groups, but it was undirected comparison. To find out how to choose TKIs or chemotherapy in different selections, TKIs versus chemotherapy was analyzed.
Four trials compared TKIs with chemotherapy in the first line setting for unselected NSCLC patients. The control arm involved standard chemotherapeutic regimens such as paclitaxel plus carboplatin or gemcitabine plus carboplatin, vinorelbine, or docetaxel. The ORR for chemotherapy was superior to the ORR for TKIs. The OS and especially the PFS tended to benefit from chemotherapy. The HR of the ORR was 3.52, and the PFS 1.29, OS 1.35 ------ all well above 1.0. These results indicate that EGFR-TKIs may be harmful to unselected patients in the first-line setting. In this situation, chemotherapy should still be used as the standard treatment for unselected patients.
Five trials reported the results of TKI treatment versus chemotherapy in patients with an EGFR mutation, showing that these patients benefitted more from TKI treatment based on both the response and PFS. The results of these five trials confirm the superiority of TKIs compared with chemotherapy for patients with an EGFR mutation. The OS for patients with an EGFR mutation was similar to chemotherapy, but there are too many factor will interfere the survival. The interference of the survival not only on the patients received TKIs in the second line but also the different chemotherapy regimen or some patients without second line therapy. The results of the meta-analysis confirmed the survival of the TKIs is similar with the chemotherapy in patients with EGFR mutations while the ORR and PFS of TKIs were dramatically superior. Patients should receive TKI treatment when their EGFR mutation status is confirmed.
In clinical practice, however, not all patients can undergo EGFR gene analysis. Salto-Tellez et al. estimated that the proportion of non-squamous NSCLC patients tested for EGFR mutations varied from 30% to 80% in East Asia. Furthermore, some patients with poor performance status or with comorbidities cannot tolerate chemotherapy. Determining whether EGFR-TKIs should be used in the first-line setting for patients with unknown EGFR status and a performance status >2 is an important issue. Two trials compared TKIs and chemotherapy in patients selected according to specific clinical characteristics. Both studies were performed in East Asia and enrolled patients with an adenocarcinoma subtype and who had never smoked or were only former light smokers. A meta-analysis of the two trials showed that the patients chosen in this way had a greater response and better PFS with TKI treatment than with chemotherapy. In the TOPICAL trial, when all NSCLC patients with poor performance status or who were unfit for chemotherapy were given erlotinib or placebo in the first-line setting, the OS did not improve with erlotinib; however, an analysis by gender showed that the female patients benefited from erlotinib in terms of PFS and OS, while the male patients did not. Thus, in some special situations involving patients with unknown EGFR status or who are not fit for chemotherapy, a trial of EGFR-TKI may be reasonable according to clinical criteria, provided that the patient and family members have been informed of the possible worsening of symptoms and disease based on the IPASS and TORCH findings.
This meta-analysis showed that groups of advanced NSCLC patients selected according to different criteria will show differential benefits from TKIs compared with standard chemotherapy. However, this meta-analysis did not consider involving individual patient data, and only two trials performed in East Asia focused on clinical selection of patients. Most trials only report the clinical characteristics and total response of the EGFR mutation and wild type patients. The different clinical factors may have different influence on the results while this meta-analysis cannot interpret. More trials are needed involving patients with specific clinical characteristics to confirm these findings.