Critical appraisal:Shen WY, Ji J, Zuo YS, Pu J, Xu YM, Zong CD, et al 2013

From Cancer Guidelines Wiki

Critical Appraisal

Article being appraised

Shen WY, Ji J, Zuo YS, Pu J, Xu YM, Zong CD, et al. Comparison of efficacy for postoperative chemotherapy and concurrent radiochemotherapy in patients with IIIA-pN2 non-small cell lung cancer: An early closed randomized controlled trial. Radiother Oncol 2013 Oct 31 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24183868.


Applicable clinical question

Key Facts

Study Design

randomised controlled trial

Number of Patients:

140


Includes an economic evaluation

no

Evidence ratings

Level of evidence

II

Risk of bias
High risk of bias Comments: This multicentre Chinese study randomised good performance status patients who had completely resected stage IIIA-pN2 NSCLC to receive 4 cycles of postoperative cisplatin/paclitaxel chemotherapy (POCT arm) or 3D conformal radiotherapy (50.4Gy) concurrently with the first 2 cycles of this chemotherapy (POCRT arm). Patients with FEV1 ≥ 1.0 L were allowed into the study. Treatments were to be administered 6-8 weeks after surgery.

The study was slow to accrue patients and closed early. In all, 140 patients were recruited to the study and 135 were randomised. The target sample size was 300 to detect an estimated (at 5 years) improvement in median survival from 30 to 40 months. Assumptions were based on a two-sided significance level of 0.05, power of 0.8, 5 years of recruitment, and 5 years additional follow-up. A CONSORT diagram indicated the disposition of all patients enrolled into the study. Of the 66 patients assigned to POCRT, 44 completed treatment, and of the 69 patients allocated to POCT, 55 completed treatment, although this difference was not statistically significant (p=0.087). Treatment discontinuations were reported mainly because of patient choice although severe oesophagitis and grade 4 myelosuppression were also mentioned. Grade 3 or 4 cytopenias occurred in approximately two-thirds of patients in each treatment arm. Grade 3 and 4 oesophagitis occurred in 8 patients and 1 patient, respectively, in the POCRT arm. One septic death occurred in the in the POCRT arm but there were no deaths in the POCT arm. Patients were stratified according to histology (squamous or non-squamous), weight loss in 6 months before study enrolment (≤ 5% or >5%), number of pN2 metastases (1 or ≥ 2), mediastinal lymph node dissection type (sampling or systematic dissection), type of surgery (pneumonectomy or lobectomy), and T stage (T1, T2, or T3). The proportion of patients who underwent mediastinal lymph node sampling was 60%. The primary study endpoints were overall survival (OS) and disease free survival (DFS). Patients were followed up for a median time of 45 months. The authors declared no conflicts of interest. A blinded trial design is not feasible when one treatment arm includes preoperative chemotherapy. However, it was unclear as to whether the treatment allocation schedule was concealed. In summary, the reliable assessment of the measurement outcomes was limited by a sample size smaller than was planned and a follow-up period that was too short.

Risk of bias assessment: randomised controlled trial

Was the trial double-blinded?
Outcomes not blinded, substantial side-effects, or not reported.
Was the treatment allocation schedule concealed?
No concealment or unclear (e.g. no approach described, open randomisation lists, person doing recruitment tossing a coin).
Were all randomised participants included in the analysis?
No exclusions or survival analysis used with all subjects included (>95% follow-up for all groups).
The field below is not considered when calculating the risk of bias rating
How was the allocation schedule generated?
Inadequate or not reported
Size of effect
4 Reason for decision: The range of estimates defined by the confidence intervals includes clinically important effects but the range of estimates defined by the confidence intervals is also compatible with no effect or a harmful effect. The primary study endpoints were overall survival (OS) and disease free survival (DFS). Median OS in the POCRT and POCT arms was 40 months (95% CI 32.9-47.7 months) and 28 months (95% CI 16.4-39.6 months), respectively. The HR reported for 5-year OS in the POCRT arm was 0.69 (95% CI 0.457-1.044; p=0.073). Median DFS in the POCRT arm was 28 months (95% CI 21.8-34.2 months) and in the POCT arm, it was 18 months (95% CI 12.2-23.8 months). The HR reported for 5-year DFS in the POCT arm was 1.49 (95% CI 1.008-2.204).
Relevance of evidence
5 Additional comments: Effects on unproven surrogate outcomes were observed. Improved DFS was observed among patients receiving POCRT and may be related to a beneficial effect of POCRT (cf. POCT) in patients with the poor prognostic feature of multiple involved mediastinal lymph nodes. A multivariate analysis of the prespecified prognostic factors showed that the only independent adverse prognostic factor was 2 or more pN2 lymph nodes. In a prespecified subgroup analysis, POCRT significantly increased the OS of patients with 2 or more pN2 lymph nodes (p=0.021).
Result of appraisal

Jutta's tick icon.png Included




Completed by

Professor Michael Brown MBBS PhD FRACP FRCPA


Jutta's tick icon.png This appraisal has been completed.


Article
Shen WY, Ji J, Zuo YS, Pu J, Xu YM, Zong CD, et al. Comparison of efficacy for postoperative chemotherapy and concurrent radiochemotherapy in patients with IIIA-pN2 non-small cell lung cancer: An early closed randomized controlled trial. Radiother Oncol 2013 Oct 31 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24183868.
Assigned to
User:Michael.brown
Topic area
Guidelines:Lung cancer/Treatment/Non-small-cell stage III operable
Clinical question
Form
Form:Critical appraisal


Section below only relevant for Cancer Council Project Officer

Edit appraisal assignment