识别与儿童、青少年和年轻成人癌症诊断相关的早期症状：一项基于人群的嵌套病例对照研究

Identifying early symptoms associated with a diagnosis of childhood, adolescent and young adult cancers: a population-based nested case-control study

DOI 原文链接

用sci-hub下载

BRITISH JOURNAL OF CANCER

影响因子:6.8

分区:医学2区 Top / 肿瘤学2区

发表日期:2024 Oct

作者: D Saatci, J Oke, A Harnden, J Hippisley-Cox

DOI: 10.1038/s41416-024-02786-5

摘要

儿童、青少年及年轻成人（CTYA，0-24岁）癌症罕见且多样，导致早期诊断面临挑战。本研究旨在探讨与后续癌症诊断相关的症状及其组合，并明确其时间框架。我们利用QResearch数据库，进行匹配的嵌套病例对照研究。采用条件逻辑回归分析在初级保健中遇到的预设症状与后续癌症诊断的关联。中位诊断间隔用于将症状划分为“晚期”和“早期”时间段，从而识别相关的早期症状。在3,186个病例和50,576个对照中，从34,247,71个CTYA人群中鉴定出。本研究发现12个新颖的关联，其中偏瘫 [OR 90.9 (95% CI 24.7-335.1)，PPV=1.6%]、阴囊肿胀 [OR 186.7 (95% CI 86.1-404.8)，PPV=2.4%] 和器官肿大 [OR 221.6 (95% CI 28.3-1735.9)，PPV=5.4%]具有显著的阳性预测值（PPV）。肢体疼痛作为儿童严重疾病的标志，是多种癌症亚型中的早期反复出现的症状。不同儿童和青少年癌症中观察到相似的临床表现。本研究利用最大规模队列，提供了症状在CTYA癌症诊断中随时间变化的预测效能的最新信息。我们的结果将有助于提高临床和公众对症状的认知，识别高风险群体，并实现更早的诊断。

Abstract

Childhood, teenage and young adult (CTYA, 0-24 years) cancers are rare and diverse, making timely diagnosis challenging. We aim to explore symptoms and symptom combinations associated with a subsequent cancer diagnosis and to establish their timeframe.Using the QResearch Database, we carried out a matched nested case-control study. Associations between pre-specified symptoms encountered in primary care and a subsequent diagnosis of any cancer were explored using conditional logistic regression. Median diagnostic intervals were used to split symptoms into "late" and "early" timeframes to identify relevant early symptoms.3186 cases and 50,576 controls were identified from a cohort of 3,424,771 CTYA. We identified 12 novel associations, of which hemiparesis [OR 90.9 (95%CI 24.7-335.1), PPV = 1.6%], testicular swelling [OR 186.7 (95%CI 86.1-404.8), PPV = 2.4%] and organomegaly [OR 221.6 (95%CI 28.3-1735.9), PPV = 5.4%] had significant positive predictive values (PPV). Limb pain, a known marker of serious illness in children, was a recurrent early symptom across cancer subtypes. Similar clinical presentations were observed across childhood and TYA cancers.Using the largest cohort to date, we provide novel information on the time-varying predictive utility of symptoms in the diagnosis of CTYA cancers. Our findings will help to raise clinical and public awareness of symptoms, stratify those at higher-risk and ultimately aid earlier diagnosis.