Two nomograms constructed for predicting the efficacy and prognosis of advanced non‑small cell lung cancer patients treated with anti‑PD‑1 inhibitors based on the absolute counts of lymphocyte subsets

Abstract

Background: Patients treated with immune checkpoint inhibitors (ICIs) are at risk of considerable adverse events, and the ongoing struggle is to accurately identify the subset of patients who will benefit. Lymphocyte subsets play a pivotal role in the antitumor response, this study attempted to combine the absolute counts of lymphocyte subsets (ACLS) with the clinicopathological parameters to construct nomograms to accurately predict the prognosis of advanced non-small cell lung cancer (aNSCLC) patients treated with anti-PD-1 inhibitors.

Methods: This retrospective study included a training cohort (n = 200) and validation cohort (n = 100) with aNSCLC patients treated with anti-PD-1 inhibitors. Logistic and Cox regression were conducted to identify factors associated with efficacy and progression-free survival (PFS) respectively. Nomograms were built based on independent influencing factors, and assessed by the concordance index (C-index), calibration curve and receiver operating characteristic (ROC) curve.

Result: In training cohort, lower baseline absolute counts of CD3⁺ (P < 0.001) and CD4⁺ (P < 0.001) were associated with for poorer efficacy. Hepatic metastases (P = 0.019) and lower baseline absolute counts of CD3⁺ (P < 0.001), CD4⁺ (P < 0.001), CD8⁺ (P < 0.001), and B cells (P = 0.042) were associated with shorter PFS. Two nomograms to predict efficacy at 6-week after treatment and PFS at 4-, 8- and 12-months were constructed, and validated in validation cohort. The area under the ROC curve (AUC-ROC) of nomogram to predict response was 0.908 in training cohort and 0.984 in validation cohort. The C-index of nomogram to predict PFS was 0.825 in training cohort and 0.832 in validation cohort. AUC-ROC illustrated the nomograms had excellent discriminative ability. Calibration curves showed a superior consistence between the nomogram predicted probability and actual observation.

Conclusion: We constructed two nomogram based on ACLS to help clinicians screen of patients with possible benefit and make individualized treatment decisions by accurately predicting efficacy and PFS for advanced NSCLC patient treated with anti-PD-1 inhibitors.

Keywords: Absolute count; Anti-PD-1inhibitor; Lymphocyte subsets; Nomogram; Non-small cell lung cancer.

Fig. 1

Patient screening flow chart

Fig. 2

The comparison of percentage and absolute counts of lymphocyte subsets between response (PR) and nonresponse (SD + PD) groups in training and validation cohorts. The comparison of absolute count of CD3⁺ (A), CD4⁺ (B), CD8⁺ (C), B (D), NK (E) cells and the percentage of lymphocyte subsets (F) between response and nonresponse group in the training cohort. The comparison of absolute count of CD3⁺ (G), CD4⁺ (H), CD8⁺ (I), B (J), NK (K) cells and the percentage of lymphocyte subsets (L) between response and nonresponse groups in the validation cohort. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Fig. 3

The relationship between ACLS and PFS in the training and validation cohorts. P-values were calculated by the log-rank test. The Kaplan–Meier analysis of the absolute counts of (A) CD3⁺, (B) CD4⁺, (C) CD8⁺, (D) B, and (E) NK cells in training cohort (n = 200). The Kaplan–Meier analysis of the absolute counts of (F) CD3⁺, (G) CD4⁺, (H) CD8⁺ AC, (I) B, and (J) NK cells in validation cohort (n = 100)

Fig. 3

The relationship between ACLS and PFS in the training and validation cohorts. P-values were calculated by the log-rank test. The Kaplan–Meier analysis of the absolute counts of (A) CD3⁺, (B) CD4⁺, (C) CD8⁺, (D) B, and (E) NK cells in training cohort (n = 200). The Kaplan–Meier analysis of the absolute counts of (F) CD3⁺, (G) CD4⁺, (H) CD8⁺ AC, (I) B, and (J) NK cells in validation cohort (n = 100)

Fig. 4

Forest plot of factors influencing the efficacy of aNSCLC patients. HR > 1 indicates that the variable is considered as a risk factor. HR < 1 represents that the variable is considered as a protective factor

Fig. 5

Forest plot of factors influencing PFS of aNSCLC patients. *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 6

The nomograms for predicting efficacy and PFS in advanced NSCLC patients. A Nomogram was developed based on two factors including baseline AC of CD3⁺ and CD4⁺ to predict probability of therapeutic response at 6-weeks after treatment. B Nomogram was developed based on six factors including hepatic metastases, tumor size, and baseline AC of CD3⁺, CD4⁺, CD8⁺ to predict probability of PFS at 4-, 8- and 12-months

Fig. 7

ROC curves and calibration curves of the nomograms for predicting efficacy and PFS in the training and validation cohorts. A ROC curves of the nomograms to predict therapeutic response at 6-weeks in both training and validation cohort. B ROC curves of the nomogram for predicting 4-, 8-, and 12-months PFS in the training and validation cohorts. C Calibration curves of nomogram for predicting therapeutic response at 6-weeks in training and validation cohorts. D Calibration curves of nomogram for predicting 4-, 8-, and 12-months PFS in training and validation cohorts

Fig. 7

ROC curves and calibration curves of the nomograms for predicting efficacy and PFS in the training and validation cohorts. A ROC curves of the nomograms to predict therapeutic response at 6-weeks in both training and validation cohort. B ROC curves of the nomogram for predicting 4-, 8-, and 12-months PFS in the training and validation cohorts. C Calibration curves of nomogram for predicting therapeutic response at 6-weeks in training and validation cohorts. D Calibration curves of nomogram for predicting 4-, 8-, and 12-months PFS in training and validation cohorts