Anomogram to predict the cumulative risk for brain metastases in patients with limited-stage small cell lung cancer without prophylactic cranial irradiation.

Abstract

Patients with small cell lung cancer (SCLC) have ahigh risk of developing brain metastases (BM). Prophylactic cranial irradiation (PCI) is astandard therapy for limited-stage SCLC (LS-SCLC) patients who achieved complete or partial response after thoracic chemoradiotherapy (Chemo-RT). Recent studies have indicated that asubgroup of patients with alower risk of BM can avoid PCI, and the present study therefore tries to construct anomogram to predict the cumulative risk of development of BM in LS-SCLC patients without PCI. After screening of 2298 SCLC patients who were treated at the Zhejiang Cancer Hospital from December 2009 to April 2016, atotal of 167consecutive patients with LS-SCLC who received thoracic Chemo-RT without PCI were retrospectively analyzed. The paper analyzed clinical and laboratory factors that may be correlated with BM, such as response to treatment, pretreatment serum neuron-specific enolase (NSE) and lactate dehydrogenase (LDH) levels, and TNM stage. Thereafter, anomogram was constructed to predict 3‑ and 5‑year intracranial progression-free survival (IPFS). Of 167patients with LS-SCLC, 50developed subsequent BM. Univariate analysis showed that pretreatment LDH (pre-LDH) ≥ 200 IU/L, an incomplete response to initial chemoradiation, and UICC stageIII were positively correlated to ahigher risk of BM (p < 0.05). Multivariate analysis identified pretreatment LDH level (hazard ratio [HR] 1.90, 95% confidence interval [CI] 1.08-3.34, p = 0.026), response to chemoradiation (HR 1.87, 95% CI 1.04-3.34, p = 0.035), and UICC stage (HR 6.67, 95% CI 1.03-49.15, p = 0.043) as independent predictors for the development of BM. Anomogram model was then established, and areas under the curve of 3‑year and 5‑year IPFS were 0.72 and 0.67, respectively. The present study has developed an innovative tool that is able to predict the individual cumulative risk for development of BM in LS-SCLC patients without PCI, which is beneficial for providing personalized risk estimates and facilitating the decision to perform PCI.