Elsevier

The Lancet

Volume 369, Issue 9577, 9–15 June 2007, Pages 1929-1937
The Lancet

Fast track — Articles
Preoperative chemotherapy in patients with resectable non-small cell lung cancer: results of the MRC LU22/NVALT 2/EORTC 08012 multicentre randomised trial and update of systematic review

https://doi.org/10.1016/S0140-6736(07)60714-4Get rights and content

Summary

Background

Although surgery offers the best chance of cure for patients with non-small cell lung cancer (NSCLC), the overall 5-year survival rate is modest, and improvements are urgently needed. In the 1990s, much interest was generated from two small trials that reported striking results with neo-adjuvant chemotherapy, and therefore our intergroup randomised trial was designed to investigate whether, in patients with operable non-small cell lung cancer of any stage, outcomes could be improved by giving platinum-based chemotherapy before surgery.

Methods

Patients were randomised to receive either surgery alone (S), or three cycles of platinum-based chemotherapy followed by surgery (CT-S). Before randomisation, clinicians chose the chemotherapy that would be given from a list of six standard regimens. The primary outcome measure was overall survival, which was analysed on an intention-to-treat basis. This study is registered as an International Standard Randomised Controlled Trial, number ISRCTN25582437.

Results

519 patients were randomised (S: 261, CT-S: 258) from 70 centres in the UK, Netherlands, Germany, and Belgium. Most (61%) were clinical stage I, with 31% stage II, and 7% stage III. Neo-adjuvant chemotherapy was feasible (75% of patients received all three cycles of chemotherapy), resulted in a good response rate (49% [95% CI 43%–55%]) and down-staging in 31% (25%–37%) of patients, and did not alter the type or completeness of the surgery (lobectomy: S: 56%, CT-S: 60%, complete resection: S: 80%, CT-S: 82%). Post-operative complications were not increased in the CT-S group, and no impairment of quality of life was observed. However, there was no evidence of a benefit in terms of overall survival (hazard ratio [HR] 1·02, 95% CI 0·80–1·31, p=0·86). Updating the systematic review by addition of the present result suggests a 12% relative survival benefit with the addition of neoadjuvant chemotherapy (1507 patients, HR 0·88, 95% CI 0·76–1·01, p=0·07), equivalent to an absolute improvement in survival of 5% at 5 years

Interpretation

Although there was no evidence of a difference in overall survival with neo-adjuvant chemotherapy, the result is statistically consistent with previous trials, and therefore adds considerable weight to the current evidence.

Introduction

Surgery is the treatment of choice for patients with disease stage I and II non-small cell lung cancer (NSCLC), but overall 5-year survival rates have remained at about 55% for stage I, 25% for stage II, and 20% for stage IIIA for several decades.1, 2 An NSCLC meta-analysis3 combined the results from eight randomised trials of surgery versus surgery plus adjuvant cisplatin-based chemotherapy and showed a small, but not significant (p=0·08), absolute survival benefit of around 5% at 5 years (from 50% to 55%). At the time the current trial was designed (mid-1990s), adjuvant chemotherapy had not become standard clinical practice, and there was also much interest in neo-adjuvant chemotherapy, which was being investigated not only in NSCLC but also in breast and gastric cancer. The clinical rationale for neo-adjuvant chemotherapy is three-fold: regression of the primary cancer could be achieved thereby facilitating and simplifying or reducing subsequent surgery; undetected micro-metastases could be dealt with at the start of treatment; and there might be inhibition of the putative stimulus to residual cancer by growth factors released by surgery and by subsequent wound healing.4, 5

In resectable stage IIIA disease, two small randomised studies6, 7 had compared surgery with or without neo-adjuvant chemotherapy, with strikingly positive results in favour of neo-adjuvant chemotherapy. However, both trials stopped early, and in each case patients randomised to the control arm (surgery alone) did not do as well as would have been expected. The benefit of neo-adjuvant chemotherapy was therefore probably over-estimated, but there was a clear indication that further research was needed to clarify its role in NSCLC.

In keeping with current practice, any new trial should include all resectable patients (given the suboptimum surgical survival rates for all stages of NSCLC), should allow a choice of chemotherapy (from a short menu of widely-used regimens), and include an assessment of quality of life (to inform future decisionmaking).

The current trial was therefore set up to compare, in patients with resectable NSCLC, surgery alone (S) versus three cycles of platinum-based chemotherapy followed by surgery (CT-S) in terms of overall survival, quality of life, pathological staging, resectability rates, extent of surgery, and time to and site of relapse.

Eligible patients had previously untreated, histologically or cytologically proven NSCLC, considered resectable by the local multidisciplinary team in accordance with local institutional policies. Patients could be any age, with WHO performance status 0–2,8 and no evidence of distant metastases. Patients had to be deemed fit for chemotherapy and the proposed surgical resection, and have no other disease or previous malignancy likely to interfere with the protocol treatments or comparisons. Local ethics committee approval of the protocol and written consent from the individual patients were required.

Patients were randomised, stratifying for participating centre, histology, WHO performance status, and clinical T and N stage, to receive either S or CT-S. To ensure balance, before randomisation and on a patient-by-patient basis, clinicians had to state which chemotherapy regimen they would use if their patient was allocated CT-S.

Patients allocated to receive CT-S could receive one of six platinum-based regimens (panel): MVP (mitomycin, vinblastine, and cisplatin); MIC (mitomycin, ifosfamide, and cisplatin); NP (vinorelbine and cisplatin); PacCarbo (paclitaxel and carboplatin); GemCis (gemcitabine and cisplatin); or DocCarbo (docetaxel and carboplatin).

Three cycles of chemotherapy were specified, with the recommendation that chemotherapy start as soon as possible after randomisation. Detailed dose modifications for each regimen were listed in the protocol, but in general patients only received a cycle of chemotherapy if they had an adequate blood count (total WBC >3000/μL, neutrophils >1500/μL, and platelets >100 000/μL) and there was no clinical evidence of infection; and for the cisplatin-based regimens no renal toxicity, ototoxicity, or neuropathy. In the management of myelotoxicity, the recommendation was that chemotherapy be delayed rather than drug dosages reduced. If radiographic changes indicated tumour progression prior to any cycle of chemotherapy, the recommendation was that patients discontinue chemotherapy and be considered for immediate surgery.

Surgery was recommended to take place as soon as possible after randomisation (for the S patients) and between 4 and 6 weeks from day 1 of the last chemotherapy cycle for CT-S patients. The surgical technique was decided by the local surgeon according to the site and extent of the tumour, and no uniform policy on nodal assessment was mandated.

Post-operative radiotherapy was not recommended.9 However, on progression, or in the case of patients who at surgery were found to be inoperable, treatment of the patient was left to the discretion of the local clinician.

Local research staff completed reports to cover the pre-randomisation assessment, chemotherapy (if allocated), surgery, pathology, and follow-up (at 1 month post-surgery, 4 and 6 months post-randomisation, 3-monthly to 2 years, and then 6-monthly). Reports included details of treatment and adverse effects, blood counts, and other relevant tests; tumour response (using the WHO definition), performance status, and details of symptoms or adverse events.

The quality of life assessments were designed to indicate any difference in patients' wellbeing by use of the SF-36 generic health status questionnaire.10 This questionnaire was selected because it assesses the effect of symptoms on daily activities and emotional distress, and there are extensive normative data available for comparison. Patients were asked to complete the SF-36 pre-randomisation, at 6 months and 1 year from randomisation, and then every year to 5 years.

The primary outcome measure was overall survival; secondary outcome measures were quality of life, clinical and pathological staging, resectability rates, extent of surgery, and time to and sites of relapse or recurrence.

With an estimated 3-year survival rate of 40% in the control (S) group, the trial was designed to detect an improvement of 15% (to 55%) in the experimental (CT-S) group, with 5% significance and 90% power, which required 450 patients and 233 events (deaths) to be seen in the two groups combined.

All randomised patients contributed to the time-to-event analyses (the intention-to-treat principle). Duration of survival was calculated from the date of randomisation to the date of death from any cause, survivors being censored at the date they were last known to be alive. Progression-free survival was calculated from the date of randomisation to the date of first progression, recurrence, or death, whichever was the sooner. Patients alive and progression-free were censored at the date they were last known to be alive and free of progression. The log-rank test was applied to compare the Kaplan-Meier curves for overall survival. The standard χ2 test for frequency was used to compare pathological staging, resectability rates, and sites of relapse.

The items of the SF-36 questionnaire were collated into eight subscales10 and the scores for these subscales were examined for changes over time and differences between the treatment arms, using a t test. The physical and mental subscales (PCS and MCS), each consisting of four of the above subscales were also compared.

Throughout, significance has been defined using the conventional 5% level.

In early 2004, we assessed that the required number of events (deaths) would not be achieved in a timely fashion, probably as a result of the case-mix of patients. Therefore the target accrual was raised to 600 patients, so that the required number of events would be achieved earlier. However, a closing accrual date of July, 2005, was set, as adjuvant chemotherapy became more standard, making randomisation to a trial with a surgery alone arm more difficult, although the overall target of 233 events for the analysis remained.

The Medical Research Council reviewed, approved, and funded this trial, and appointed an independently led Trial Steering Committee to oversee its execution. An independent Data Monitoring Committee reviewed the interim data at regular, usually annually, meetings. No formal stopping rules were used. The data analysis, preparation of the manuscript, and interpretation were done independently from the funding source. The trial results have been reported according to CONSORT guidelines. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Section snippets

Results

Between July, 1997 and July, 2005, 519 (S: 261, CT-S: 258) patients were randomised from 70 centres in the UK (342 patients), Netherlands (144 patients), Germany (22 patients), and Belgium (11 patients; figure 1). The median age of patients was 63 years, 374 (72%) were men, 283 (55%) had WHO performance status 0, 256 (49%) had squamous cell histology, and the two groups were well matched at baseline (table 1).

Most patients were staged by bronchoscopy and CT of the thorax. Further investigations

Discussion

This large intergroup trial, which accrued over 500 patients through a successful collaboration between the MRC, NVALT, and the EORTC, compared surgery alone with three cycles of platinum-based chemotherapy followed by surgery. The underlying rationale was that neo-adjuvant chemotherapy would downstage tumours, deal with micro-metastases, and inhibit the surgical stimulus to any residual cancer, and as a result, based on positive results for neo-adjuvant chemotherapy in several small trials,

References (18)

There are more references available in the full text version of this article.

Cited by (0)

View full text