|Concomitant radio-chemotherapy based on platin compounds in patients with locally advanced non-small cell lung cancer (NSCLC): a meta-analysis of individual data from 1764 patients
|Auperin A, Le Pechoux C, Pignon J P, Koning C, Jeremic B, Clamon G, Einhorn L, Ball D, Trovo M G, Groen H J, Bonner J A, Le Chevalier T, Arriagada R
This was a well-conducted and reported individual patient data meta-analysis. It concluded that concomitant platin-based radio-chemotherapy may improve survival of patients with locally advanced non-small cell lung cancer, compared with radiotherapy alone, but further research is needed. The authors' cautious conclusions appear reliable.
To investigate whether platin-based concomitant chemotherapy might improve survival in patients with locally advanced non-small cell lung cancer, using individual patient data meta-analysis.
MEDLINE, EMBASE, CANCERLIT, the Physician Data Query (PDQ) clinical trial registry and the Cochrane Library were searched from 1985 to 2002. Search terms were reported. No language restrictions were applied. Manual searches of reference lists, review articles, books and meeting proceedings of the American Society of Clinical Oncology and the International Association for the Study of Lung Cancer were also performed. Trial investigators and clinical experts were asked to help locate trials.
Randomised controlled trials (RCTs) comparing radiotherapy combined with concomitant cisplatin or carboplatin chemotherapy to radiotherapy alone, as treatment for patients with locally advanced unresectable or inoperable non-small cell lung cancer without distant metastasis, were eligible for inclusion. Trials had to be of patients who had not received prior radiotherapy or chemotherapy. Only trials with adequate methods of randomisation, and where the radiotherapy regimen was the same in both groups, were included. Trials that finished recruiting after the year 2000 were excluded.
Radiotherapy schedules varied between trials for total dose, dose per fraction number of daily doses, and treatment duration (two courses of two weeks at a time, to a continuous course of seven weeks). Chemotherapy was cisplatin, carboplatin or etoposide combined with cisplatin or carboplatin, and given daily, weekly, or every two, three or four weeks. The treatment and control arms appeared well-balanced for patient characteristics: 78% of patients were male; median age was 61 (range 31 to 83 years); 44% had performance status of zero and 52% of one; 57% had squamous carcinoma and the most common stage was IIIa (60% of patients). Median follow-up ranged from four to 15.8 years.
The authors did not stated how studies were selected for the review.
Assessment of study quality
Data were checked for internal consistency and against published results. Any queries were amended after discussion with the trial investigators.
Basic patient demographic and disease status data, as well as date of randomisation, dates of first recurrence and death were requested from the trial investigators. The primary outcome was overall survival, which was calculated from the date of randomisation to the date of death from any cause. Patients still alive were censored at their final follow-up assessment. The secondary outcome was event-free survival which was time to the first event (death or recurrence). The length of follow-up was estimated by the reverse Kaplan-Meier method.
Methods of synthesis
Hazard ratios (HR) and 95% confidence intervals (CI) were calculated for each trial individually and overall in a pooled analysis using a fixed-effect model. All analyses were on an intention to treat basis. χ2 tests and the I2 statistic were used to assess heterogeneity. A Cox proportional hazards model was also used, which was stratified by trial. Absolute differences in two and five year survival were calculated using the pooled hazard ratio and the relevant survival rates in the control arms. Kaplan-Meier curves were plotted (using all data with no stratification by trial). Sensitivity analyses were used to assess the impact of small trials, old trials and trials with incomplete data. Indirect comparisons were used to assess if the treatment effects varied according to the type and administration methods of the chemotherapy and the radiotherapy. Interaction terms were added to the model to assess the impact of treatment in the following pre-defined subgroups: gender, age (<61, 61 to 70, >70 years), performance status, stage and histopathology.
Results of the review
Data from nine trials (n=1,764 patients, n=959 patients for radiotherapy and chemotherapy, n=805 patients for radiotherapy only) were included. It was not possible to obtain data from two small trials and not all data were available in another trial, so these three were excluded from the meta-analysis.
Survival: Overall there were 1,657 deaths (92%). Radiotherapy and chemotherapy significantly improved survival compared with radiotherapy alone, hazard ratio 0.89 (95% Confidence Interval (CI): 0.81 to 0.98; p=0.02) which corresponds to increases in two-year survival from 21.4% to 25.4% and for five-year survival from 6.0% to 8.2%. There was a low level of heterogeneity (I2=32%). Sensitivity analyses showed that when two small trials were removed there was still a marginally significant (p=0.059) reduction in death rates but that when the two oldest trials, or the two trials with incomplete data were removed, the observed statistically significant benefit of radiotherapy and chemotherapy was no longer observed.
Event-free survival: Data on the first recurrence (metastases or loco-regional) were available for 1,072 patients (61%). Radiotherapy and chemotherapy significantly improved event-free survival compared with radiotherapy alone, hazard ratio 0.84 (95% CI: 0.74 to 0.96; p=0.009). This corresponds to increases in two-year event-free survival from 21.3% to 27.3% and for five-year event-free survival from 6.4% to 9.9%. There was moderate heterogeneity (I2=55%). When the two oldest trials, or the two trials with incomplete data were removed, the observed statistically significant benefit or radiotherapy and chemotherapy was no longer observed.
Indirect and subgroup analyses: Most of the indirect comparisons did not show any statistically significant differences between different chemotherapy or radiotherapy regimens. Single agent cisplatin or carboplatin showed shorter survival than a single agent with etoposide combination (hazard ratio of 0.93 compared with 0.72; p=0.05). Radiotherapy and chemotherapy showed better performance in patients with stage IIIb disease compared with stage IIIa disease for survival (hazard ratio 0.81 compared with 1.01; p=0.053) and event-free survival (hazard ratio 0.76 compared with 1.02; p=0.047). Radiotherapy and chemotherapy was also more effective depending on patient age, appearing to improve survival more in older patients (hazard ratio 0.67 for over 70; 0.80 for ages 61 to 60; 1.08 for under 61; p=0.001).
Concomitant platin-based radio-chemotherapy may improve the survival of patients with locally advanced non-small cell lung cancer but the available data was insufficient to reliably confirm this. Further research is needed to define the size of any treatment benefit and the optimal chemotherapy schedule.
This was an individual patient data meta-analysis conducted by a dedicated meta-analysis group. The inclusion criteria were clearly specified and the search appeared to be thorough and appropriate. Every effort was made to locate relevant trials, including contact with trial investigators and experts. Raw data were re-checked and results compared against the published trial results. The statistical analysis methods were appropriate, subgroup analyses were pre-specified and sensitivity analyses were performed to investigate sources of between study heterogeneity. The results were clearly presented and their relationship with other similar research was discussed. Limitations of the available evidence and of the results found in this meta-analysis were also considered. The cautious conclusions of this well-conducted and reported meta-analysis appear reliable.
Implications of the review for practice and research
Practice: The authors did not state any recommendations for practice.
Research: The authors plan to update a previous meta-analysis on locally advanced non-small cell lung cancer to estimate the effect of sequential chemotherapy, look at the impact of new agents such as taxanes, and directly compare concomitant and sequential treatment.
Auperin A, Le Pechoux C, Pignon J P, Koning C, Jeremic B, Clamon G, Einhorn L, Ball D, Trovo M G, Groen H J, Bonner J A, Le Chevalier T, Arriagada R. Concomitant radio-chemotherapy based on platin compounds in patients with locally advanced non-small cell lung cancer (NSCLC): a meta-analysis of individual data from 1764 patients. Annals of Oncology 2006; 17(3): 473-483
Subject indexing assigned by NLM
Aged; Antineoplastic Agents /adverse effects /therapeutic use; Carcinoma, Non-Small-Cell Lung /drug therapy /radiotherapy; Combined Modality Therapy; Disease-Free Survival; Female; Humans; Lung Neoplasms /drug therapy /radiotherapy; Male; Middle Aged; Platinum Compounds /adverse effects /therapeutic use
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.