Survival outcomes following complete mediastinal lymphadenectomy or selective mediastinal lymphadenectomy in patients with stage I–IIIA non-small cell lung cancer: protocol for a systematic review and meta-analysis

Abstract

Introduction

Lung cancer remains the largest cause of cancer-related deaths worldwide. Surgical removal of non-small cell lung cancer (NSCLC) has the potential to achieve a cure, although there is ongoing debate regarding the significance of removing mediastinal nodes and the optimal extent of lymph node excision. The purpose of this research is to assess the survival outcomes in patients diagnosed with stage I–IIIA NSCLC who received either complete mediastinal lymphadenectomy (CML) or selective mediastinal lymphadenectomy (SML).

Methods and analysis

The protocol follows the guidelines recommended in Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols, and this meta-analysis will be conducted in accordance with the standard methodology recommended by the Cochrane Collaboration and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidance. We will conduct a comprehensive search for randomised controlled trials and non-randomised studies examining the effectiveness of CML compared with SML in patients with stage I–IIIA NSCLC. Two authors will perform a comprehensive search of the MEDLINE/PubMed, Embase, the Cochrane Library, CNKI, WanFang, Sinomed, VIP and Web of Science databases. There will be no restrictions on language or publication date, and the search will be conducted on 10 April 2024, with ongoing searches for new research. Reference lists will also be checked and pertinent journals will be hand searched. The primary outcomes include overall survival (OS) and disease-free survival (DFS), while the secondary outcomes consist of 1-year, 3-year and 5-year OS rates and 1-year, 3-year and 5-year DFS rates. Two independent reviewers will screen, extract data, assess quality and evaluate the potential for bias in the selected research, with a third acting as arbitrator. Subgroup analyses and sensitivity analyses are planned. The quality of the evidence will be evaluated using Grading of Recommendations Assessment, Development and Evaluation. Review Manager V.5.4 will be used for the analysis and synthesis process.

Ethics and dissemination

Ethical review and approval are not necessary for this study because it is based on a secondary analysis of the literature. The results will be submitted for reporting in a peer-reviewed publication.

Study registration

Open Science Framework (https://doi.org/10.17605/OSF.IO/PN7UQ).

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The study will be carried out according to the recommendation of the Cochrane Handbook for Systematic Reviews of Interventions.

• Two researchers will independently screen studies, extract data, assess quality and determine levels of confidence, and a third researcher will arbitrate any disagreements.

• This systemic review protocol follows the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols guidelines and the results will be reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidance.

• The main potential limitation of the planned analysis could be the potential heterogeneity among studies included in the analysis.

• The meta-analysis is dependent on the data reported by the primary investigators, and the calibre of the original studies will impact the overall quality of the evidence.

Introduction

In 2023, the American Cancer Society predicts that there will be approximately 127 070 deaths related to lung cancer and 238 340 new cases of the disease in the USA.¹ Lung cancer is still the leading cause of cancer-related mortality worldwide.^{2 3} Approximately 80% of histological variations of lung cancer fall under the category of non-small cell lung cancer (NSCLC), with lung adenocarcinoma and lung squamous cell carcinoma being the predominant subtypes.^{4 5}

Early detection of lung cancer, accurate staging and appropriate treatment interventions are crucial to improving patients’ survival prospects. Surgical removal of NSCLC has the potential to achieve a cure, although there is ongoing debate regarding the significance of removing mediastinal nodes and the optimal extent of lymph node excision.^6–13 Researches have demonstrated that early micrometastatic tumour cell dissemination to local lymph nodes happens in certain patients with NSCLC.^14–17 The pathological state of patients with lung cancer can be precisely staged with the use of intraoperative excision of lymph nodes and surrounding tissues in pertinent places.¹⁸ This serves as a foundation for later adjuvant treatment, thereby enhancing patient survival. Systematic lymph node dissection, however, may have disadvantages such as a broad surgical scope, excessive trauma and an increase in postoperative problems, which may lower patients’ quality of life following surgery.¹⁹ Moreover, the incidence of lymph node metastasis is not very high for some patients with early-stage NSCLC, and the majority of patients may not have regional lymph node metastases.

According to three prior meta-analyses,^20–22 complete mediastinal lymphadenectomy (CML) increases patients’ long-term survival in stage I–IIIA NSCLC, but Mokhles et al ²⁰ were very cautious about the reliability of such a conclusion. Nonetheless, the outcomes of our earlier meta-analysis demonstrated that in patients with early-stage NSCLC, CML and selective mediastinal lymphadenectomy (SML) had comparable overall survival (OS),²³ and other studies have also shown that lymph node dissection performed in early-stage lung cancer does not affect patient survival.^24–27 Therefore, newer systematic reviews and meta-analyses are needed to reconcile these discrepancies, and conclusive evaluations can offer more compelling justifications for the selection of a certain course of treatment. For these reasons, we will carry out a systematic review and meta-analysis of existing randomised controlled trials (RCTs) and non-randomised studies to evaluate the survival outcome of CML versus SML in patients with stage I–IIIA NSCLC.

Methods and analysis

Protocol and registration

This systematic review protocol has been registered at the Open Science Framework (https://doi.org/10.17605/OSF.IO/PN7UQ). The protocol follows the guidelines recommended in Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) (PRISMA-P checklist in online supplemental file 1).²⁸ The meta-analysis will be conducted in accordance with the standard methodology recommended by the Cochrane Collaboration and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidance.^29–31

Inclusion criteria

The population, intervention, comparators, outcomes and study design framework³² is adhered to by the inclusion criteria. All studies that fulfil the prerequisites for inclusion will be accepted without regard to language or year of publication.

1. Population. Patients with stage I–IIIA NSCLC who underwent either CML or SML will be included without restrictions on country, race, ethnic origin, age, sex or occupation.

2. Intervention: CML.

3. Comparators: SML.

4. Outcomes. The main measures include OS and disease-free survival (DFS), while the additional measures consist of 1-year, 3-year and 5-year OS rates and 1-year, 3-year and 5-year DFS rates.

5. Study design. RCTs as well as non-randomised studies will be incorporated into the research.

Exclusion criteria

• Duplicate texts and articles.

• There are obvious errors in the data results.

• Relevant and usable data cannot be obtained from the literature; full-text articles cannot be obtained after exhaustive searches.

• Case studies, conference abstracts, review articles, letters, editorials and expert opinions without original data.

Data sources and search strategy

In this systematic review, a comprehensive search will be conducted in the following electronic bibliographical databases: PubMed, Embase, the Cochrane Library, CNKI, WanFang, Sinomed, VIP and Web of Science. There will be no restrictions on the language or date of publication, and the search will be conducted on 10 April 2024, with ongoing searches for new research. The full search strategy is presented in online supplemental file 2. All of the retrieved papers’ reference lists will be examined and searched to find additional research that might be pertinent. Unpublished research will be found by searching the World Lung Cancer Conference and the American Society of Clinical Oncology. Additionally, reference lists of pertinent review articles will be examined to identify any potentially overlooked literature or grey literature, and pertinent journals will be hand searched.

We will strive to contact significant authors in order to gather thorough information for eligible studies. We will request any relevant details, including additional materials that may not have been fully disclosed or published, as well as data from unofficial sources related to relevant research.

Study selection

There are two steps involved in the process of selecting a study. Duplicates of the papers will be eliminated in the first stage. Following this, the titles and abstracts of the documents will be reviewed to assess their potential relevance. In the second step, received documents will be appraised, whether they are deemed uncertain or potentially relevant. The second phase then involves a thorough analysis of the entire texts of the documents. The reasons for the exclusion of any studies during this step will be described. Lastly, the PRISMA flow diagram (figure 1) presents the study selection procedure. If two reviewers have differing opinions, they will consult a third reviewer and make a final decision together.

Figure 1.

Flow diagram showing the selection process for inclusion. *Consider, if feasible to do so, reporting the number of records identified from each database or register searched (rather than the total number across all databases/registers). **If automation tools were used, indicate how many records were excluded by a human and how many were excluded by automation tools.

Data extraction

The data will be extracted by two authors separately using a defined procedure. Any discrepancies between the two reviewers will be resolved through discussion or by consulting the third author. The collected data included several baseline characteristics, and the characteristics of studies are attached as table 1. If there are missing or unclear data in a research paper, the authors will be contacted to obtain more information. The HRs for time-to-event data will be obtained either directly from the original studies or estimated by interpreting survival curves, as recommended by Parmar et al. ¹³

Table 1.

General information of the included studies

First author, year
Country
Design
Sample size (males/females)
Age
Outcome diagnostic criteria
Inclusion or exclusion
Reason(s) for exclusion
Follow-up (mean and range) (months)
Tumour size (cm)
Tumour stage
Surgery procedure
Adjuvant treatment
Primary outcomes: Overall survival Disease-free survival
Second outcomes: 1-year, 3-year and 5-year overall survival rates 1-year, 3-year and 5-year disease-free survival rates

Methodological quality assessment

Two researchers will evaluate the possibility of bias in the selected studies by employing the guidelines outlined in the Cochrane Handbook for Systematic Reviews of Interventions.³³ The evaluation criteria will focus on seven areas, including random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting and other biases.³⁴ The quality of evidence for all RCTs will be evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) method.³⁵

We will assess the methodological rigour of non-randomised studies by employing the Newcastle–Ottawa Scale.³⁶ This tool assesses the methodological quality of non-randomised controlled clinical studies based on the following three major components: selection of the study groups, comparability of the study groups and ascertainment of the exposure of interest in cohort study or the outcome of interest in case–control study, with a maximum score of 9. The articles will be evaluated for potential bias according to their overall scores. A score of 0–3 will denote low quality, while a score of 4–6 will indicate moderate quality and a score of 7–9 will signify high quality.³⁷

Two independent assessors will evaluate the quality. Any discrepancies between the two assessors will be addressed through deliberation and consensus with a senior researcher. The final conclusions will be reviewed by two senior researchers.

Data synthesis and statistical analysis

The statistical analysis will be performed using Review Manager V.5.4 software, developed by the Cochrane Collaboration in London, UK. This software enables users to input protocols, conduct comprehensive reviews, and incorporate textual content, study characteristics, comparison tables and study data, as well as conduct meta-analyses. Meta-analysis will be performed using random-effects or fixed-effects methods, depending on the presence or absence of significant heterogeneity. Heterogeneity among the studies will be assessed using the χ² test and I² statistic.³⁸ For the χ² statistic, a p value of <0.10 will be considered statistically significant for heterogeneity; for the I² statistic, heterogeneity was interpreted as absent (I²: 0–25%), low (I²: 25–50%), moderate (I²: 50–75%) or high (I²: 75–100%).^{38 39} When the presence of heterogeneity is established, the random-effects approach will be employed.^{40 41} In the event of non-significant heterogeneity, the fixed-effects method will be used to amalgamate the findings.^{42 43}

The comparison of time-to-event outcomes will be conducted through the use of HR, while dichotomous data will be compared using risk ratio or OR. The respective 95% CIs will be computed for each estimate and displayed in forest plots. All statistical assessments were two sided, and a p value of <0.05 will be considered to indicate statistical significance.

Subgroup analysis

When variability is identified, subgroup analysis will be performed to examine the distinct characteristics of the studies included in order to ascertain the underlying cause of the variability.

Sensitivity analysis

We will perform sensitivity analyses using the leave-one-out approach to validate the precision and dependability of the meta-analysis outcomes. These analyses will assess the impact of each individual study on the overall conclusions, providing a more comprehensive evaluation of the meta-analysis results’ reliability.

Assessment of publication biases

If the meta-analysis incorporates 10 or more studies in its findings, the detection of publication bias is evaluated by employing a funnel plot and the Egger’s test. Asymmetry in the funnel plot indicates the presence of publication bias and/or a potential small study effect.^{44 45}

Grading quality of evidence

Two evaluators will appraise the level of certainty of evidence using the GRADE framework.⁴⁶ This framework is widely embraced for evaluating the credibility of evidence and the robustness of recommendations. The results will be classified as high, moderate, low or very low quality, taking into account four key factors: risk of bias, inconsistency, indirectness and imprecision. High-quality findings will indicate a high level of confidence in the effectiveness and quality of the intervention.^47–49 The evaluations conducted using the GRADE framework will be presented in a tabular format for concise summarisation.

Patient and public involvement

None.

Ethics and dissemination

Ethical review and approval are not necessary for this study because it is based on a secondary analysis of the literature. The results will be submitted for reporting in a peer-reviewed publication.

Discussion

Over the past decade, numerous studies have been published on the topic of CML versus SML for NSCLC.^50–54 Currently, the surgical excision of tumours and the exploration of lymph nodes continue to be a feasible strategy for managing NSCLC. However, the comprehensive removal of lymph nodes entails a significant surgical procedure, substantial physical trauma, a higher chance of post-surgery complications and death, and potentially reduced quality of life. Hence, there remains considerable controversy regarding the necessity of performing an extensive lymph node dissection during clinical operations and practice for individuals diagnosed with early-stage NSCLC.^6–13 In North America, surgeons demonstrate a greater tendency to conduct mediastinal staging through surgical means before the operation. They are also less inclined to recommend surgical intervention when N2 disease is detected before the operation, and are more prone to using induction therapy before resection. On the other hand, European surgeons may propose surgical intervention as the primary course of treatment, followed by supplementary therapy for specific instances of N2 disease. Additionally, they may conduct a more assertive intraoperative examination of the lymph nodes.⁵⁵

Debates persist about whether CML or SML is more necessary for treating early-stage NSCLC. For these reasons, our study aims to conduct a meta-analysis to evaluate the survival outcome of CML and SML in treating patients with early-stage NSCLC. The goal is to compare the survival advantages of these two surgical techniques and lymph node dissection in the treatment of early-stage NSCLC, in order to provide guidance for selecting appropriate surgical approaches. The meta-analyses have several advantages. All evidence syntheses will adhere to the previously described methods in a standardised manner.

However, there are some potential limitations. The meta-analysis is dependent on the data reported by the primary investigators, and the calibre of the original studies will impact the overall quality of the evidence. Our research is based on existing data, and there is a potential for bias in published studies. We will carefully assess the presence of publication bias whenever there are enough studies available. The main potential limitation of the planned analysis could be the potential for heterogeneity among studies included in the analysis. Factors such as differences in ethnic makeup, various treatment methods, varying lengths of follow-up and discrepancies in the number of participants lost to follow-up could limit the conclusions drawn from this meta-analysis. To address this issue, we will make every effort to locate reliable research and overcome this constraint by conducting subgroup analysis.

In summary, this meta-analysis seeks to use the results of previous literature research to offer proactive advice for clinical practice. The research is anticipated to help compare the survival advantages of these two surgical methods for treating early-stage NSCLC. Consequently, it is expected that the findings will guide the preferred treatment approach in medical literature, ultimately reducing mortality rates for patients undergoing surgery for stage I–IIIA NSCLC.

Ethics statements

Patient consent for publication

Not applicable.