Development and Validation of a Prediction Formula for Survival Time in Patients with Lung Squamous Cell Carcinoma

Background Lung cancer has become the leading cause of cancer-related mortality. The 5-year overall survival rate of lung squamous cell carcinoma (LUSC) remains less than 15%. Identifying patients who are likely to experience short or long survival has clinical utility by helping to minimize overtreatment or undertreatment. However, there is currently no method to predict survival time based on available information. Aim To develop and validate a prediction formula for survival after surgery for LUSC using readily available biomarkers. Methods The inclusion criteria of patients were male, who was diagnosed LUSC underwent radical surgery in two hospitals between April 2015 and December 2018. The methods consist of three parts as pooled analysis, phase 1 and 2. Before analyzing the two phases, a pooled analysis was performed to determine whether data from two hospitals could be combined and analyzed together. In phase 1, the prediction formula was developed using biomarkers that were proved reliable. In phase 2, the validation test was conducted to verify the accuracy of the developed formula.Patient data were as follows: 1. Demographics, 2. Laboratory data of Complete blood count (CBC) and biochemistry values were measured prior to the LUSC radical operation, 3. Biomarkers included Advanced Lung Cancer Inflammation Index (ALI), Hemoglobin, Alb, Lymphocyte, Platelets, Neutrophil/Lymphocyte Ratio and Systemic immune-inflammation index, with hemoglobin concentration (Hb). 4. Outcome measure was 5-year survival, defined as survival status during 5 years after radical surgery. All data were compared between two groups divided by the survival status at 5 years after the operation. Results 69 patients with LUSC were enrolled. 1. In phase 1, The prediction formula was developed as follows: Survival time = [6.9 × ALI] + [93.4 × Hb] -198.6 (days). 2. In phase 2, Validation test using 39 enrolled LUSC patients showed that the error rate was − 4% (SD 28%). Conclusion The prediction formula of survival time in patients with squamous cell lung cancer was developed using ALI and Hb as follows: Survival time (days) = [6.9 × ALI] + [93.4 × Hb] -198.6. The ROC curve analysis showed that the cut-off value of ALI was 48.04 (AUC = 0.690, p = 0.007). The novel prediction formula was developed and seemed feasible with an error rate of -4% (standard deviation of 28%), indicating that the newly developed prediction formula seems to be clinically feasible.