Prognostic Role of Neutrophil-Lymphocyte Ratio
This analysis was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. An electronic search of the following databases was undertaken: Medline (host: OVID) from 1946 to January 2013; EMBASE (host: OVID) from 1974 to January 2013; Cochrane Database of Systematic Reviews from 2005 to November 2012; American Society of Clinical Oncology abstracts 2011 to 2013; and European Society of Medical Oncology abstracts 2011 to 2012. (It was expected that data presented earlier would be captured in full publications.) Search terms included "cancer," "neutrophils," "lymphocytes," and "ratio." Citation lists of retrieved articles were screened manually to ensure sensitivity of the search strategy. The full search strategy is described in the Supplementary Methods (available online) http://jnci.oxfordjournals.org/content/106/6/dju124/suppl/DC1.
Inclusion criteria for the primary analysis were as follows: 1) studies of people with solid tumors reporting on the prognostic impact of the peripheral blood NLR, and 2) availability of a hazard ratio (HR) and 95% confidence interval (CI) or a P value for overall survival (OS). For a secondary analysis, studies providing a hazard ratio for cancer-specific survival (CSS), progression-free survival (PFS), disease-free survival (DFS), or recurrence-free survival (RFS) were included as well. Duplicate publications were excluded. Two reviewers (A. Templeton, M. McNamara) evaluated independently all of the titles identified by the search strategy. The results were then pooled, and all potentially relevant publications were retrieved in full. The same two reviewers then assessed the full articles for eligibility. Inter-reviewer agreement was assessed using Cohen's kappa. Disagreement was resolved by consensus. Corresponding authors were contacted to clarify missing or ambiguous data. To avoid inclusion of duplicated or overlapping data, we compared author names and institutions where patients were recruited and contacted authors to address potential concerns. In cases where no answer was obtained and substantial doubts remained, the study reporting fewer patients was not included in the analysis.
OS was the primary outcome of interest. CSS, PFS, and DFS were secondary outcomes. Data were collected using predesigned abstraction forms. The following details were extracted: name of first author, type of publication (abstract, full text), year of publication, journal, number of patients included in analysis, disease site, disease stage (nonmetastatic, metastatic, mixed [nonmetastatic and metastatic]), collection of data (prospective, retrospective), cutoff defining high NLR used for peripheral blood NLR, consideration of receiver operating characteristic curves (C-index) for selection of cutoff where available, and hazard ratios and associated 95% confidence intervals for OS, PFS, DFS, or RFS as applicable. Hazard ratios were extracted preferentially from multivariable analyses where available. Otherwise, hazard ratios from univariate analyses were extracted.
The meta-analysis was conducted initially for all included studies for each of the endpoints of interest. Subgroup analyses were conducted for predefined parameters such as disease site, disease stage, whether data were derived from univariate or multivariable analyses, and whether data were published in abstract form or as full articles. Disease site subgroups were generated for the main outcome if at least three studies on that site were available; the remaining studies were pooled in a subgroup termed "other."
Extracted data were combined into a meta-analysis using RevMan 5.1 analysis software (Cochrane Collaboration, Copenhagen, Denmark). Estimates of hazard ratios were weighted and pooled using the generic inverse-variance and random-effect model. Analyses were conducted for all studies, and differences between the subgroups were assessed using methods described by Deeks et al.. Meta-regression analysis was performed to evaluate the effect of NLR cutoff on the hazard ratio for OS. Publication bias was assessed by visual inspection of the funnel plot. Heterogeneity was assessed using Cochran Q and I statistics. All statistical tests were two-sided, and statistical significance was defined as P less than .05. No correction was made for multiple testing.
Methods
Data Sources and Searches
This analysis was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. An electronic search of the following databases was undertaken: Medline (host: OVID) from 1946 to January 2013; EMBASE (host: OVID) from 1974 to January 2013; Cochrane Database of Systematic Reviews from 2005 to November 2012; American Society of Clinical Oncology abstracts 2011 to 2013; and European Society of Medical Oncology abstracts 2011 to 2012. (It was expected that data presented earlier would be captured in full publications.) Search terms included "cancer," "neutrophils," "lymphocytes," and "ratio." Citation lists of retrieved articles were screened manually to ensure sensitivity of the search strategy. The full search strategy is described in the Supplementary Methods (available online) http://jnci.oxfordjournals.org/content/106/6/dju124/suppl/DC1.
Study Selection
Inclusion criteria for the primary analysis were as follows: 1) studies of people with solid tumors reporting on the prognostic impact of the peripheral blood NLR, and 2) availability of a hazard ratio (HR) and 95% confidence interval (CI) or a P value for overall survival (OS). For a secondary analysis, studies providing a hazard ratio for cancer-specific survival (CSS), progression-free survival (PFS), disease-free survival (DFS), or recurrence-free survival (RFS) were included as well. Duplicate publications were excluded. Two reviewers (A. Templeton, M. McNamara) evaluated independently all of the titles identified by the search strategy. The results were then pooled, and all potentially relevant publications were retrieved in full. The same two reviewers then assessed the full articles for eligibility. Inter-reviewer agreement was assessed using Cohen's kappa. Disagreement was resolved by consensus. Corresponding authors were contacted to clarify missing or ambiguous data. To avoid inclusion of duplicated or overlapping data, we compared author names and institutions where patients were recruited and contacted authors to address potential concerns. In cases where no answer was obtained and substantial doubts remained, the study reporting fewer patients was not included in the analysis.
Data Extraction
OS was the primary outcome of interest. CSS, PFS, and DFS were secondary outcomes. Data were collected using predesigned abstraction forms. The following details were extracted: name of first author, type of publication (abstract, full text), year of publication, journal, number of patients included in analysis, disease site, disease stage (nonmetastatic, metastatic, mixed [nonmetastatic and metastatic]), collection of data (prospective, retrospective), cutoff defining high NLR used for peripheral blood NLR, consideration of receiver operating characteristic curves (C-index) for selection of cutoff where available, and hazard ratios and associated 95% confidence intervals for OS, PFS, DFS, or RFS as applicable. Hazard ratios were extracted preferentially from multivariable analyses where available. Otherwise, hazard ratios from univariate analyses were extracted.
Data Synthesis
The meta-analysis was conducted initially for all included studies for each of the endpoints of interest. Subgroup analyses were conducted for predefined parameters such as disease site, disease stage, whether data were derived from univariate or multivariable analyses, and whether data were published in abstract form or as full articles. Disease site subgroups were generated for the main outcome if at least three studies on that site were available; the remaining studies were pooled in a subgroup termed "other."
Statistical Analyses
Extracted data were combined into a meta-analysis using RevMan 5.1 analysis software (Cochrane Collaboration, Copenhagen, Denmark). Estimates of hazard ratios were weighted and pooled using the generic inverse-variance and random-effect model. Analyses were conducted for all studies, and differences between the subgroups were assessed using methods described by Deeks et al.. Meta-regression analysis was performed to evaluate the effect of NLR cutoff on the hazard ratio for OS. Publication bias was assessed by visual inspection of the funnel plot. Heterogeneity was assessed using Cochran Q and I statistics. All statistical tests were two-sided, and statistical significance was defined as P less than .05. No correction was made for multiple testing.
SHARE
![Pediatric Supportive Care (PDQ®): Supportive care - Patient Information [NCI]-Overview](https://cdnus.oss-us-west-1.aliyuncs.com/libs/images/NewsJpg/cate_6/072_215x160.jpg)
