Yang 2014.
| Methods | Design: parallel‐group Randomisation: yes, method stated Blinding: open‐label Withdrawals: stated |
|
| Participants | Setting: multicentre study, hospital outpatient department Number eligible: 253 Number enrolled: 236 Number in treatment group: 118 Number in control group: 118 Number of withdrawals (treatment/control): 4/0 Number completing trial (treatment/control): 12/46 Age range: treatment 24 to 81 years; control 31 to 79 years Sex: 59 M, 177 F Ethnicity: East Asian NSCLC diagnosis: histologic/cytologic diagnosis of non‐squamous NSCLC, stage IIIB to IV disease Inclusion criteria: chemotherapy‐naive patients of East Asian ethnicity and unknown EGFR mutation status. Stage IIIB to IV non‐squamous NSCLC. Age ≥ 18 years, "light ex smokers" or "never smokers" measurable disease by RECIST version 1.0, ECOG PS 0 or 1 Exclusion criteria: known EGFR status before study entry, documented brain metastasis (previously treated stable brain metastases were allowed), clinically significant third space fluid collections, inability to interrupt aspirin or other non‐steroidal anti‐inflammatory agents (except aspirin at a dose of 1300 mg daily for a 5‐day period) and concomitant use of CYP3A4 inducers Baseline characteristics of treatment/control groups: comparable |
|
| Interventions |
PC/Gefitinib arm Pemetrexed (500 mg/m2) + cisplatin (75 mg/m2) on day 1 of 21‐day cycle. Maximum of 6 cycles. Then non‐progressing patients received gefitinib 250 mg daily as maintenance Gefitinib arm Gefitinib 250 mg daily as maintenance |
|
| Outcomes | Progression‐free survival Overall survival Tumour response – RECIST Time to progressive disease (TtPD) Duration of response (DoR) ASEs – NCI‐CTC Association between EGFR mutation status and clinical outcomes |
|
| Notes | — | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | "... randomisation was controlled by a centrally located computerised voice response unit using a computer‐generated random sequence and an interactive voice response system..." Comment: this was judged as a low risk of bias |
| Allocation concealment (selection bias) | Low risk | External computer generated random sequence Comment: this was judged as a low risk of bias |
| Blinding (performance bias and detection bias) All outcomes | Low risk | Open‐label but review authors judge that outcome is not likely to be influenced by lack of blinding Comment: this was judged as a low risk of bias |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Withdrawals presented in Figure 1. 58 patients completed the study, with balanced numbers between both arms. Comment: this was judged as a low risk of bias |
| Selective reporting (reporting bias) | Low risk | All prespecified outcomes were reported Comment: this was judged as a low risk of bias |
| Other bias | Unclear risk | Authors have declared paid consultancies, honorarium and research funding from industry Comment: this was judged as an unclear risk of bias |