Fig. 2
Factors selection and nomogram construction based on peripheral-blood inflammatory-nutritional indices of PFS. a The LASSO coefficient profiles of candidate variables; b the optimal tuning parameter lambda λ in LASSO analysis selected with tenfold cross-validation. Dotted vertical line is set at the nonzero coefficients selected via tenfold cross-validation. The first vertical line equals the minimum error. The second vertical line shows cross-validated error within one standard error of the minimum; c forest plot of multivariate analysis of PFS. d “ESCLL” nomogram (based on ECOG PS, preSII, changeCAR, changeLYM and postLDH) was established to assess 1-year and 2-year PFS; Each factor corresponds to a specific point by drawing a vertical line from that variable to the points axis. After sum of the scores for each variable located on the Total Points axis. Finally, the sum represents the probability of 1-, 2-year survival by drawing straight down to the survival axis. (For example, an aNSCLC patient with changeLYM ≥ − 0.20, preSII < 625, postLDH ≥ 228, ECOG PS < 2 and changeCAR ≥ 0.38, the total score will be given by 249, corresponding to 1- and 2-year risks of progression of 0.645 and 0.868, respectively. The patient will accordingly have approximately 35.5% and 13.2% survival probabilities at 1 and 2 years, respectively) *P < 0.05; **P < 0.01; ***P < 0.001. e The ROC curve of the nomogram for 1-year, 2-year PFS; f calibration curves of the nomogram for 1-year, 2-year PFS; g decision curve analysis of the nomogram for 1-year, 2-year PFS; h Kaplan–Meier curves depicting PFS according to risk levels (cutpoint value: 1.37). Significant difference in survival of patients was observed between high and low risk score. ECOG PS, Eastern Cooperative Oncology Group Performance Status scores; preSII, pre-treatment systemic immune inflammation index; postLDH, post-treatment lactate dehydrogenase; changeLYM, change of absolute lymphocyte count; changeCAR, change of C-reactive protein-albumin ratio; ROC, receiver operator characteristic