Neutrophil to lymphocyte with monocyte to lymphocyte ratio and white blood cell count in prediction of lung cancer
Thang Thanh Phan, An Thi Thuy Nguyen, Anh Ngoc Van Nguyen, Hang Thuy Nguyen, Toan Trong Ho, Suong Phuoc Pho, Binh Thanh Mai, Son Truong Nguyen
Abstract
Background
Lung cancer is the most common cause of cancer deaths in both sexes, while it is very difficult for screenings and early detection.
Aims
This study aims to clarify the role of systematic inflammation markers, including white blood cell (WBC), neutrophil (NEU), monocyte (MONO), platelet (PLT), neutrophil to lymphocyte ratio (NLR), monocyte to lymphocyte ratio (MLR) and platelet to lymphocyte ratio (PLR) in prediction of lung cancer.
Methods
A case-control study was conducted on 1,315 primary lung cancer patients and 1,315 healthy adults with matched age and gender at Cho Ray hospital. NLR, MLR and PLR were calculated by using neutrophil, lymphocyte, monocyte and platelet count which were recalled from laboratory database. With 600 cases in the derivation set, the logistic regression with univariate analysis was used to identify the impacted marker, then developing the optimal prediction model for lung cancer by logistic regression with multivariate method. The diagnostic values of optimal model consisting of sensitivity (Sen), specificity (Spe), positive predictive value (PPV), negative predictive value (NPV) and the area under the ROC curve (AUC) value were extracted and verified on all data, in validation set.
Results
The median values of WBC, NEU, MONO, PLT, NLR, MLR and PLR in lung cancer were not significantly difference between histological subtypes and clinical stages (p > 0.05), but higher than the values in control group (p < 0.01). Multivariates analysis shows that NLR, MLR and WBC were three parameters that have the significant impact of the optimal prediction model (p < 0.01). The AUC value, sensitivity and specificity of the optimal model for lung cancer detection were 0.881, 73.5 per cent (95 per cent CI:70.3–76.6) and 87.7 per cent (95 per centCI:85.2–89.9), respectively. Whereas, the PPV and NPV values of prediction model were 85.7 per cent (95 per cent CI:82.8–88.2) and 76.8 (95 per centCI:73.9–79.5), respectively. Among three biomarkers, the AUC values of NLR (0.853) and MLR (0.842) were higher than the value of WBC (0.752) (p < 0.01).
Conclusion
The results of this study show that NLR with MLR and WBC in optimal prediction model are promising biomarkers for lung cancer screening that could be applied in clinical practice with the advantage of convenience and low cost.
Full Text:
Lung cancer is the most common cause of cancer deaths in both sexes, while it is very difficult for screenings and early detection.
Aims
This study aims to clarify the role of systematic inflammation markers, including white blood cell (WBC), neutrophil (NEU), monocyte (MONO), platelet (PLT), neutrophil to lymphocyte ratio (NLR), monocyte to lymphocyte ratio (MLR) and platelet to lymphocyte ratio (PLR) in prediction of lung cancer.
Methods
A case-control study was conducted on 1,315 primary lung cancer patients and 1,315 healthy adults with matched age and gender at Cho Ray hospital. NLR, MLR and PLR were calculated by using neutrophil, lymphocyte, monocyte and platelet count which were recalled from laboratory database. With 600 cases in the derivation set, the logistic regression with univariate analysis was used to identify the impacted marker, then developing the optimal prediction model for lung cancer by logistic regression with multivariate method. The diagnostic values of optimal model consisting of sensitivity (Sen), specificity (Spe), positive predictive value (PPV), negative predictive value (NPV) and the area under the ROC curve (AUC) value were extracted and verified on all data, in validation set.
Results
The median values of WBC, NEU, MONO, PLT, NLR, MLR and PLR in lung cancer were not significantly difference between histological subtypes and clinical stages (p > 0.05), but higher than the values in control group (p < 0.01). Multivariates analysis shows that NLR, MLR and WBC were three parameters that have the significant impact of the optimal prediction model (p < 0.01). The AUC value, sensitivity and specificity of the optimal model for lung cancer detection were 0.881, 73.5 per cent (95 per cent CI:70.3–76.6) and 87.7 per cent (95 per centCI:85.2–89.9), respectively. Whereas, the PPV and NPV values of prediction model were 85.7 per cent (95 per cent CI:82.8–88.2) and 76.8 (95 per centCI:73.9–79.5), respectively. Among three biomarkers, the AUC values of NLR (0.853) and MLR (0.842) were higher than the value of WBC (0.752) (p < 0.01).
Conclusion
The results of this study show that NLR with MLR and WBC in optimal prediction model are promising biomarkers for lung cancer screening that could be applied in clinical practice with the advantage of convenience and low cost.
PDF