条件相对生存率（CRS）对向患者传达预后非常有用，因为它提供了在治疗后存活一段时间后的预期寿命估计。本研究估计了食管或胃癌患者在存活一定时期后的三年相对生存率。从荷兰癌症登记处选择了2006-2020年被治愈意图治疗（手术切除伴随或不伴随（新）辅助治疗，或化疗放疗）的非转移性食管或胃癌患者。自切除或放疗的最后一天开始计算CRS。如果患者在诊断后存活1、3和5年，则存活另外三年的概率（即三年CRS）为食管和胃癌分别为62％，79％，87％和69％，84％，90％。稍微存活了3年后，接受食管癌切除术的患者（n = 12,204）的三年CRS分别为0、I、II和III病理分期的91％，88％，77％和60％，接受化疗放疗的患者（n = 4,158）的三年CRS在临床II和III病理分期分别为51％和66％。稍微存活了3年后，接受胃癌切除术的患者（n = 6531）的三年CRS分别为0、I、II和III病理分期的99％，90％，73％和59％。尽管食管或胃癌患者的预后不良，但在患者治疗后存活了几年后，寿命显著增加。本研究为在治疗后的随访期间向患者传达预后提供了有价值的信息。本文受版权保护，保留所有权利。

Conditional relative survival (CRS) is useful for communicating prognosis to patients as it provides an estimate of the life expectancy after having survived a certain time after treatment. This study estimates the three year relative survival conditional on having survived a certain period for patients with esophageal or gastric cancer. Patients with non-metastatic esophageal or gastric cancer diagnosed between 2006-2020 treated with curative intent (resection with or without (neo)adjuvant therapy, or chemoradiotherapy) were selected from the Netherlands Cancer Registry. CRS was calculated since resection or last day of chemoradiotherapy. The probability of surviving an additional three years (i.e. three year CRS), if the patients survived 1,3 and 5 years after diagnosis was 62%, 79%, 87% and 69%, 84%, 90% for esophageal and gastric cancer, respectively. The three year CRS after having survived 3 years for patients with esophageal cancer who underwent a resection (n=12,204) was 91%, 88%, 77% and 60% for pathological stage 0, I, II and III, and for patients with esophageal cancer who received chemoradiotherapy (n=4,158) was 51% and 66% for clinical stage II and III, respectively. The three year CRS after having survived 3 years for patients with gastric cancer who underwent a resection (n=6,531) was 99%, 90%, 73% and 59% for pathological stage 0, I, II and III, respectively. Despite poor prognosis of patients with esophageal or gastric cancer, life expectancy increases substantially after patients have survived several years after treatment. This study provides valuable information for communication of prognosis to patients during follow-up after treatment.This article is protected by copyright. All rights reserved.