2026年9月23日，美国食品药品监督管理局（FDA）顾问小组将召开会议，就Galleri检测进行投票表决。Galleri是一种旨在同时筛查50余种癌症的血液检测。这是美国专家委员会首次在上市前批准申请程序中正式审查多癌种筛查血液检测。迄今为止，此类检测均未经过该程序便已面向患者推出，美国国家癌症研究所也明确指出，目前尚无任何多癌种检测获得FDA授权。本文将解释委员会究竟需要就哪些问题进行表决、为何投票结果并不等同于批准，以及赞成票或反对票对于正在考虑是否自费进行癌症筛查血液检测的人意味着什么。.
FDA小组的表决内容
医疗器械咨询委员会分子与临床遗传学小组将于美国东部时间上午9时至下午6时，在马里兰州银泉市FDA白橡园区召开会议。其任务明确而具体：就GRAIL公司为Galleri检测提交的上市前批准申请进行讨论、提出建议并进行投票表决。.
FDA将该产品定义为一种仅凭处方使用的、基于新一代测序技术的体外诊断检测，通过分析常规抽血所获得的游离DNA中的癌症特异性甲基化模式来筛查癌症。该检测适用于50岁及以上的成年人，旨在作为推荐筛查项目的补充，而非替代现有筛查手段。当检测到异常信号时，该检测还能预测信号可能来源于体内的哪个部位，从而为后续诊断检查指明方向。.
FDA已开放公众意见征集，编号为FDA-2026-N-8004，征集截止日期为2026年9月16日。本次会议向公众开放，并提供网络直播。.
为何顾问小组的投票结果不等于批准
这正是大多数新闻标题容易混淆的地方。咨询委员会提供的是独立专家意见，其建议不具有法律约束力。FDA通常会参考这些建议，但并无法律义务必须遵从，且FDA在此之后可能还需数周乃至数月才会作出最终决定。因此，9月23日的小组投票只是一个信号，而非最终裁决。.
还有第二个值得关注的区别。Galleri在美国作为实验室自主研发检测项目销售，该类别依据无需证明对患者具有临床获益的法规进行监管。该检测于2018年获得的突破性设备认定也并非上市前审查。因此，无论最终结果如何，此次会议是相关基础证据首次接受这种形式的公开审查。.
摆在桌面上的证据
此次申请主要依据两组数据。第一组是PATHFINDER 2，这是一项在美国开展的研究，共纳入25,490名参与者，随访期为一年。第二组是英国NHS-Galleri试验第一轮筛查的干预组数据，共纳入逾70,000名参与者，是迄今唯一一项在目标人群中开展的多癌种检测随机对照试验。.
我们此前的文章详细介绍了 多癌种早期检测的完整试验结果, ，包括NHS-Galleri主要终点未能达到的原因。下表总结了委员会需要解决的核心矛盾。.
| 专家组面临的问题 | 支持的理由 | 反对的理由 |
|---|---|---|
| 目前尚无筛查手段的癌症 | 一次抽血即可发现卵巢癌或胰腺癌的信号 | 发现癌症并不等同于挽救生命 |
| 证据质量 | NHS-Galleri是同类规模最大的随机试验 | 其关于晚期癌症的主要终点未能达到 |
| 预测受累器官 | 准确预测原发部位可缩短诊断流程 | 预测错误反而会延长诊断时间 |
| 假阳性 | 特异性较高，接近99% | 超过一半的阳性结果最终未发现癌症 |
| 可及性与费用 | 抽血比结肠镜检查更易被接受 | 自费费用约为$900，且无常规医保报销 |
最新科学进展
根据PubMed收录的研究，2026年是技术快速进步与临床问题悬而未决并存的一年，而这正是专家组所面临的局面。.
2026年9月发表于《公共卫生挑战》的一篇评论文章，超越了PATHFINDER 2的表面数据，深入探讨该检测对常规医疗实践及癌症公共卫生问题究竟能带来哪些实质贡献（DOI）。简而言之：发现更多癌症并不自动等同于帮助更多患者，作者认为这一差距必须用结局数据而非检出率来弥合。.
《癌症流行病学、生物标志物与预防》杂志上一篇方法学文章从统计学角度阐述了同样的观点：任何筛查检测的获益都取决于疾病的自然病程，即特定癌症的生长速度，以及早期发现是否真的能改变预后（DOI）。这对您意味着什么：一项检测在技术上可以非常出色，却仍然无法延长任何人的寿命——而这正是随机对照试验所要揭示的问题。.
《北美胃肠内镜诊所》2026年发表的一篇关于血液结直肠癌筛查的综述，对这类检测提出了一个实际警示：对晚期癌前病变的敏感性较低，且异常结果后接受结肠镜随访的比例令人失望（DOI）。筛查检测只有在患者采取行动时才能发挥作用。另一篇文章详细介绍了 经FDA批准的结肠癌血液检测, ，这是一项针对单一癌症的检测，其监管审批过程也有所不同。.
最后，2026年7月《肿瘤学与血液学评论》发表的一篇综述，对遗传性癌症风险人群（即额外筛查理由最为充分的群体）中的液体活检进行了研究，但仍得出结论：尚无随机对照试验证实其临床获益（DOI).
获批对您意味着什么
如果专家组投票支持，且FDA随后跟进，多项情况将同时发生变化，但没有一项是立竿见影的。.
- 监管层面。该检测将从实验室自主研发检测类别转入FDA已对其声明用途进行证据审查的监管框架。.
- 处方层面。该检测仍需凭处方使用，适用于50岁及以上的成年人，因此临床医生仍需为您开具检测并解读结果。.
- 保险覆盖层面。FDA授权并不等同于保险覆盖。医疗保险（Medicare）和商业保险公司需单独决定，这一过程有其自身的时间周期。.
- 诊疗指南层面。目前没有任何专业学会或美国预防服务工作组（USPSTF）的建议涵盖多癌种检测。FDA授权本身并不会改变诊疗指南。.
- 您现有的筛查项目。即便获批，也不意味着您可以跳过乳腺钼靶检查、结肠镜检查或宫颈癌筛查。该检测的定位是补充手段，而非替代。.
如果专家组投票反对，实际情况与今天基本相同：该检测仍可获取，未经此途径审查，且需自费支付。.
What to do while the decision is pending
The most useful step this week is also the least dramatic one: keep up the screening already proven to work for your age and risk profile. That is the recommendation the National Cancer Institute repeats regardless of what any multi-cancer test shows.
If you are considering paying for a multi-cancer test now, three questions are worth asking your doctor before you order it. What would we do with a positive result, and how many follow-up exams might that mean? Would a negative result change anything about my other screening, and the answer should be no. And who pays if the workup finds nothing?
Talk to a clinician promptly, whatever any screening test says, if you notice unexplained weight loss, blood in the stool or urine, a lump that persists, a cough that will not settle, or unusual bleeding. Symptoms are evaluated on their own merit. This guide explains 肿瘤标志物及其局限性, which are a different tool used mostly for monitoring rather than screening.
词汇表
| 学期 | 意义 |
|---|---|
| Premarket approval (PMA) | The FDA’s most demanding review pathway for a medical device or test |
| Advisory committee | A panel of outside experts whose vote advises the FDA without binding it |
| Laboratory-developed test | A test designed and run inside a single lab, under rules that do not require proof of patient benefit |
| Cell-free DNA | DNA fragments floating in the blood, some of them shed by tumor cells |
| DNA甲基化 | Small chemical tags on DNA whose pattern can point to a cancer |
| Cancer signal origin | The organ the test predicts a detected signal came from |
| 特异性 | A test’s ability to avoid alarming people who do not have cancer |
| Natural history | How a disease behaves over time if nobody intervenes |
常见问题解答
Is the Galleri test FDA approved?
Not as of September 2026. It has been offered as a laboratory-developed test, and the National Cancer Institute states that no multi-cancer detection test has been authorized by the FDA. The September 23 advisory committee meeting is a step in that review, not the end of it.
What cancers does the Galleri test look for?
It is designed to detect a shared cancer signal across more than 50 cancer types, including several with no organized screening program, such as ovarian and pancreatic cancer. It does not cover every cancer, and it is not a substitute for mammography, colonoscopy, or cervical screening.
How accurate is it?
Specificity is high, close to 99%, so false alarms are relatively rare. Sensitivity is much lower, especially for early-stage disease, which means a normal result cannot rule cancer out. In published work, more than half of people with a positive result had no cancer found after workup.
How much does it cost, and does insurance pay?
The test costs around $900 in the United States, generally paid out of pocket. No multi-cancer test is currently recommended for reimbursement by the Centers for Medicare and Medicaid Services, and follow-up diagnostic procedures may add cost that is also uncertain to be covered.
If the FDA approves it, should I get one?
That is a conversation with your doctor rather than an automatic yes. Authorization would mean the evidence passed a formal review for a stated use. It would not mean the test is recommended for everyone, covered by insurance, or a replacement for screening you already do.
What happens if my result is positive?
A positive result is a signal, not a diagnosis. It triggers further tests, usually imaging and sometimes a biopsy, often guided toward the organ the test predicted. That workup can take weeks, and in a substantial share of cases no cancer is found.
来源
- US Food and Drug Administration – September 23, 2026 Molecular and Clinical Genetics Panel meeting announcement
- US National Cancer Institute – Questions and answers about multi-cancer detection tests
- US National Cancer Institute – Tumor markers fact sheet
- PubMed: Sium and Lucero-Prisno, Public Health Challenges, 2026 (DOI); Lange and Etzioni, Cancer Epidemiology, Biomarkers and Prevention, 2026 (DOI); Xue, Daruwalla and Liang, Gastrointestinal Endoscopy Clinics of North America, 2026 (DOI); Kuhlen et al., Critical Reviews in Oncology and Hematology, 2026 (DOI).
延伸阅读
- Our earlier article covers 多癌种早期检测的完整试验结果
- 本指南为您说明 肿瘤标志物及其局限性
- Another article details 经FDA批准的结肠癌血液检测
- 我们的指南涵盖 a complete method for reading blood test results
利用 AI DiagMe 了解您的实验室结果
Whatever the FDA decides this week, most of the numbers on your lab report are the ordinary ones: blood counts, liver and kidney values, glucose, iron. AI DiagMe reads your blood, urine, or stool report in plain language and tells you what each line measures, what an out-of-range value can suggest, and what is worth raising with your doctor. It helps you understand, never to diagnose, and it does not replace a healthcare professional.
