6533b81ffe1ef96bd1277ada

RESEARCH PRODUCT

Comparison between gemcitabine-based combination (G) and single-agent chemotherapy (S) for elderly patients (EP) with advanced non-small cell lung cancer (NSCLC): A literature-based meta-analysis

subject

description

19586 Background: It was estimated that a quarter of all patients who have a diagnosis of NSCLC worldwide are more than 70 years old. This meta-analysis tries to shed light on the controversial results of phase III trials evaluating in NSCLC EP doublets against third generation S. Methods: We performed a literature search using MEDLINE and Cochrane Library. We selected only clinical trials responding to the question of our meta-analysis. Outcomes recorded were 1-year survival rate (1-y SR), overall response rate (ORR) and haematological toxicity (HT). Fixed-effects and random-effects models were used to calculate pooled odds ratios (OR). An OR greater than 1 indicates that doublet is more effective for 1-y SR and ORR and more toxic for HT. Results: Three published randomized controlled phase III trials (SICOG 9909; MILES; AISCAP-SICOG) were selected yielding a total of 1082 patients (G: 426; S: 655) clustered in seven comparisons. Drugs delivered to randomized patients included gemcitabine, vinorelbine and paclitaxel. EP treated with doublets showed respect to control patients a pooled estimate for 1-y SR advantage of 36%, not statistically significant (OR=1.356; 95% CI=0.925–1.990; p>0.05). The pooled estimate for ORR advantage was 57% and statistically significant (OR=1.559; 95% CI=1.220–2.015; p<0.05). However G showed not significant difference for HT (OR=1.168; 95% CI=0.685–1.992; p>0.05). Conclusions: These data confirm in EP superior efficacy and equal tolerability of G in comparison with S previously demonstrated for adult patients. Anyway G seems to not change prognosis of NSCLC EP. It is worthy to note that all the trials analysed showed some biases: early closure of the study, second-line therapy or crossover, lower dosage of drugs in combination regimen, inclusion of unfit or strongly comorbid patients. This meta-analysis doesn’t solve troubles in decision-making of treatment for EP, but suggest to design a better phase III trial including more patients and improving accrual criteria for reducing biases. An indication of a potentially active combination regimen (gemcitabine + paclitaxel) is suggested in SICOG 9909 trial. No significant financial relationships to disclose.