Concomitant Radio-Chemotherapy Based On Platin

introduction

Lung cancer remains a major cause of death worldwide with

over 1.1 million deaths per year [1]. Non-small cell lung

cancer (NSCLC) accounts for at least 80% of all lung tumours.

About 35% of these patients present with locally advanced

non-metastatic disease [2] for whom radical thoracic

radiotherapy is frequently part of treatment. The NSCLC

Collaborative Group meta-analysis [3] showed that sequential

cisplatin-based chemotherapy given in addition to radical

radiotherapy prolonged survival in patients with locally

advanced NSCLC. However, the prognosis for these patients

remained poor with a 3-year survival rate of approximately

14%. Cytotoxic agents used as radiosensitisers given at the

same time as radiotherapy have been evaluated against

radiotherapy alone in several randomised trials [4вЂ"14].

A primary aim of this combined approach was to improve

survival by increasing loco-regional control, while the

sequential use of chemotherapy has been directed at reducing

the rate of distant metastasis. However, the majority of the

trials performed have reported inconclusive results and it is

original

article

*Correspondence to: Dr A. AupeÐ'Ò' rin, Unit of Biostatistics and Epidemiology,

Institut Gustave-Roussy, 94805 Villejuif Cedex, France. E-mail: auperin@igr.fr

Ð'Є 2006 European Society for Medical Oncology

still controversial whether concomitant radio-chemotherapy

does in fact improve survival of patients with locally advanced

NSCLC. The size of most of these trials has not been large

enough to detect a 5 to 10% increase in survival and there

were also some heterogeneities in the trial designs.

The Meta-Analysis of Cisplatin/carboplatin based

Concomitant Chemotherapy in non-small cell Lung Cancer

(MAC3-LC) Group was therefore created to undertake

a meta-analysis based on all available individual patient data

from randomised trials. Combining the data of these trials

could provide increased statistical power, time to event

analyses conducted in intention-to-treat and subgroup

analyses leading to greater ability to determine whether

concomitant chemotherapy might lead to a moderate

improvement in survival in patients with locally

advanced NSCLC.

methods

selection criteria

Trials were eligible provided they randomised patients with locally

advanced unresectable or inoperable NSCLC without distant metastasis

to receive radiotherapy alone or radiotherapy combined with concomitant

systemic chemotherapy based on cisplatin or carboplatin. Trials had to

have been adequately randomised. Trials were considered if the

radiotherapy regimen was the same in both arms. Patients should not

have received prior radiotherapy or chemotherapy. Published and

unpublished trials without language restriction were eligible. Trials with

patient accrual completed after 2000 were ineligible.

search methods

Trials published between 1985 and 2002 were sought by searching

electronic databases (Medline, Embase, Cancerlit) without language

restriction, using the search terms: (Carcinoma, Non-Small Cell Lung/drug

therapy, lung neoplasms) and (Carcinoma, Non-Small Cell Lung/

radiotherapy, lung neoplasms) and (clinical trial phase III, randomised

controlled trials). This was supplemented by manual searches (reference

lists of trial publications, review articles, relevant books, meeting

proceedings of American Society of Clinical Oncology and International

Association for the Study of Lung Cancer). The Physician Data Query

(PDQ) clinical trial registry and the Cochrane library were also searched

and investigators and experts were asked to help identify trials.

individual patient data

Individual patient data were collected for all randomised patients.

The following data were requested: gender, age, performance status at the

time of randomisation, stage, histopathology, randomisation date,

allocated treatment, and updated information on survival and on first

recurrence. Data were checked for internal consistency and with published

results. Amendments were made as necessary through discussion with the

investigators.

statistical analysis

The main endpoint was overall survival which was evaluated from the

time of randomisation until death due to any