Two years ago, we (4), first demonstrated the usefulness of switch maintenance with an EGFR-TKI. The phase III SATURN (Sequential Tarceva UnResectable NSCLC) trial randomly assigned 889 advanced NSCLC patients without disease progression after completion of 4 cycles of standard platinum based chemotherapy to receive erlotinib or placebo. Notably, tissue collection was mandatory for enrollment. The trial met its primary end point of PFS, demonstrating a significant improvement for patients receiving erlotinib (median PFS 12.3 versus 11.1 weeks; HR 0.71, 95% CI, 0.62-0.82, P<0.0001). The co-primary end point of the study, PFS in the subgroup of EGFR immunohistochemistry positive (defined as EGFR expression on the membrane of >10% of cells) patients, was also met (HR 0.69, 95% CI, 0.58-0.82, P<0.0001). In the whole population, the PFS benefit in the active arm translated in survival benefit (median OS 12.0 versus 11.0 months; HR 0.81, 95% CI, 0.70-0.05, P=0.009). In this trial an extensive biomarkers analysis was performed, including EGFR mutational status. As expected, patients having an EGFR mutation had a significant PFS improvement (median PFS 44.6 versus 13 weeks; HR 0.10, 95% CI, 0.04-0.25, P<0.001); furthermore, also for EGFR wild type population the PFS favored the erlotinib arm (HR 0.78, 95% CI, 0.63-0.96, P=0.02).

The WJTOG0203 trial (7), randomized 604 Japanese patients with advanced NSCLC after completion of 3 cycles of platinum doublet chemotherapy, to receive three additional cycles of the same regimen or gefitinb. The study failed to meet its primary end point of overall survival. However, maintenance gefitinib significantly prolonged PFS (4.6 versus 4.3 months; HR 0.68, 95% CI, 0.57-0.80, P<0.001), with the greatest benefit observed in patients with adenocarcinoma histology (5.1 versus 4.4 months; HR 0.60, 95% CI, 0.50-0.73, P<0.001), that is the histotype classically associated with presence of EGFR mutations.

Sequential gefitinib after first line standard chemotherapy was also tested in white population enrolled in the EORTC 08021 trial (8). With the main limits of a low accrual - leading to early closure of the trial - and an ambitious statistical design - in which primary end point was to improve survival of 28% (from 11 to 14 months) - patients receiving gefitinib had longer PFS than those receiving placebo (median PFS 4.1 versus 2.9 months, HR 0.61, 95% CI, 0.45-0.83, P<0.0001), confirming the potential role of gefitinib in maintenance setting.

In a recent issue of Lancet Oncology, Zhang et al. (9) reported the final results of the INFORM (Iressa in NSCLC FOR Maintenance, C-TONG0804) trial, a phase III study of gefitinib versus placebo as maintenance treatment in Chinese patients with molecularly unselected locally advanced or metastatic NSCLC who had achieve disease control after completion of 4 cycles of platinumbased doublet chemotherapy. The study, enrolling 148 patients per arm, met its the primary end-point of PFS. Patients treated with gefitinb had a 58% relative reduction in risk of progression compared with those receiving placebo (4.8 versus 2.6 months; HR 0.42, 95% CI, 0.33-0.55, P<0.0001), while overall survival did not differ between the two groups (median survival 18.7 versus 16.9 months; HR 0.84, 95% CI, 0.62-1.14, P=0.26). Sequential anti-EGFR therapy was also associated with higher disease control rate (72% versus 51%, P=0.0001) and better symptom control. Although tissue collection was not mandatory for study entry, 79 patients (27%) provided tumor tissue for EGFR status assessment and activating mutations were found in 30 samples (38%). Compared to ITT population, in EGFR mutant patients the improvement in PFS was greater (16.6 versus 2.8 months; HR 0.17, 95% CI, 0.07-0.42, P<0.0001) with a HR quite similar to that observed in SATURN trial, with no evidence of benefit in the EGFR wild type population.

How should we interpret the INFORM data in the context of clinical practice? How gefitinib maintenance data compare to erlotinib results? Looking at the PFS curve of SATURN and INFORM, it seems that the outcome of patients included in the INFORM study was better. Clearly the difference in PFS observed between the two studies was largely influenced by the difference in study populations. In fact, the INFORM study was conducted in China, a geographic area with higher incidence of EGFR mutations when compared to western countries, while the SATURN included less than 15% of Asiatic patients. Analysis of EGFR mutated patients in the two studies showed comparable results: both erlotinib and gefitinib produced a similar PFS benefit, with approximately 90% reduction in the risk of progression. Importantly, in the EGFR wildtype population, only erlotinib produced a significant PFS improvement, confirming previous data showing that gefitinib works only in EGFR mutated while erlotinib produces some benefit, modest but statistically significant, even in absence of EGFR mutations. Probably the most interesting finding comes from survival analysis. In the INFORM study, no survival difference between gefitinib and placebo was detected, while in the SATURN trial, the modest improvement in PFS translated in a significant survival difference favoring erlotinib. Looking at the HR, in both SATURN and INFORM, the reduction in risk of death was similar (HR=0.81 in SATURN and HR=0.83 in INFORM), suggesting a marginal efficacy difference between the two drugs. Moreover, it is not possible to exclude that INFORM failed to meet the overall survival end-point because of the high percentage of patients with EGFR mutations (approximately 40%) and therefore because of the confounding effect of post-study therapies including further administration of EGFR-TKIs.

Finally, INFORM data confirmed again that EGFR mutations are the best predictor of response to an EGFRTKI and consequently EGFR mutant patients gain the greater benefit when treated early during the course of their disease. Moreover, it is confirmed that Asian patients are a “naturally enriched population” with a higher incidence of hidden EGFR mutations: In the INFORM the HR for progression in EGFR unknown individuals was 0.40, superimposable to that in the ITT population (HR=0.42) and median survival time reported in both groups as well as response rate after first-line chemotherapy (37%) are aligned with other trials conducted in Eastern countries, even if in a different setting (10,11). Furthermore also in the WJTOG0203 (7), only considering the most favorable subgroup (i.e. adenocarcinoma histology, non-smokers), median survival time was 23.5 and 25.1 months for patients in the chemotherapy arm and gefitinib arm respectively.