Stratification of Patients With Stage IB NSCLC Based on the 8th Edition of the American Joint Committee on Cancer (AJCC) Staging Manual

Abstract

Objective: To assess the postoperative prognosis of patients with stage IB non-small cell lung cancer (NSCLC), using a prognostic model (PM). Methods: Patients with stage IB of NSCLC from the two academic databases {the Surveillance, Epidemiology, and End Results [SEER-A, N = 1,746 (training cohort)], Sun Yat-sen University Cancer Center [SYSUCC, N = 247 (validation cohort)], and SEER-B (N = 1,745)} who had undergone lung surgery from 2001 to 2015 were enrolled. The primary clinical endpoint was cancer-specific survival (CSS). Covariate inclusion of prognostic indicators was carried out using a multivariable two-sided P < 0.05. We identified and integrated significant prognostic factors for survival in the training cohort to build a model that could be validated in the validation cohort. We used univariate analysis to evaluate the utilized ability of PM in the different races/ethnicities. Results: CSS discrimination in the PM was comparable in both the training and validation cohorts [C index = 0.66(SEER-A), 0.67(SYSUCC), and 0.61(SEER-B), respectively]. Discretization with a fixed PM cutoff of 291.5 determined from the training dataset yielded low- and high-risk subgroups with disparate CSS in the validation cohort (training cohort: hazard ratio [HR] 2.724, 95% confidence intervals [CI], 2.074-3.577; validation cohort: SEER-B HR 1.679, 95% CI, 1.310-2.151, SYSUCC HR 3.649, 95% CI 2.203-6.043, all P < 0.05). Our five-factor PM was able to predict CSS; 48-month CSS was 87% in the low-risk subgroup vs. 69% in the high-risk subgroup for the training cohort, while in the validation cohort, they were 80 vs. 73%(SEER-B) and 84 vs. 60% (SYSUCC), respectively. In addition, the results showed that PM with all unadjusted HR > 1 was a significant risk prognostic indictor in white men (P < 0.001), Chinese people (P < 0.001), and other races (P = 0.012). Conclusion: We established and validated a PM that may predict CSS for patients with IB NSCLC in different races/ethnicities, and thus, help clinicians screen subgroups with poor prognosis. In addition, further prospective studies and more cases from different regions are necessary to confirm our findings.

Keywords: NSCLC; prognostic model; stage IB; survival; treatment strategy.

Figure 1

Flow chart of the patient screening process in the the Surveillance, Epidemiology, and End Results.

Figure 2

The diagram of the patient screening process in the Sun Yat-sen University Cancer Center.

Figure 3

Cancer-specific survival curve for patients with stage IB NSCLC according to the prognostic model in the training cohort (A), internal validation cohort (B), and external validation cohort (C).

Figure 4

(A) Cancer-specific survival curve for NSCLC patients with stage IA, low-risk group of IB, high-risk group of IB, and stage IIA; (B) Cancer-specific survival curve for NSCLC patients with stage IA, and low-risk group of IB; (C) Cancer-specific survival curve for NSCLC patients with stage IIA, and high-risk group of IB.

Figure 5

Impact of prognostic model on survival in different races/ethnicities.

Figure 6

Cancer-specific survival curve for stage IB NSCLC according to the status of TSPI (0: negative, 1: positive).